Carbohydrate-based DNA ligands: Sugar-oligoamides as a tool to study carbohydrate-nucleic acid interactions

Article abstract of DOI:10.1021/ja050794n

Sugar-oligoamides have been designed and synthesized as structurally simple carbohydrate-based ligands to study carbohydrate-DNA interactions. The general design of the ligands 1-3 has been done as to favor the bound conformation of Distamycin-type γ-linked covalent dimers which is a hairpin conformation. Indeed, NMR analysis of the sugar-oligoamides in the free state has indicated the presence of a percentage of a hairpin conformation in aqueous solution. The DNA binding activity of compounds 1-3 was confirmed by calf thymus DNA (ct-DNA) NMR titration. Interestingly, the binding of the different sugar-oligoamides seems to be modulated by the sugar configuration. Semiquantitative structural information about the DNA ligand complexes has been derived from NMR data. A competition experiment with Netropsin suggested that the sugar-oligoamide 3 bind to DNA in the minor groove. The NMR titrations of 1-3 with poly(dA-dT) and poly(dG-dC) suggested preferential binding to the ATAT sequence. TR-NOE NMR experiments for the sugar-oligoamide 3-ct-DNA complex both in D₂O and H₂O have confirmed the complex formation and given information on the conformation of the ligand in the bound state. The data confirmed that the sugar-oligoamide ligand is a hairpin in the bound state. Even more relevant to our goal, structural information on the conformation around the N-glycosidic linkage has been accessed. Thus, the sugar asymmetric centers pointing to the NH-amide and N-methyl rims of the molecule have been characterized.

Full text of DOI:10.1021/ja050794n

Published on Web 06/08/2005
Carbohydrate-Based DNA Ligands: Sugar-Oligoamides as a
Tool to Study Carbohydrate-Nucleic Acid Interactions
Jason N. Martin,† Eva M. Mun˜oz,† Caroline Schwergold,† Florence Souard,†
Juan Luis Asensio,† Jesu´s Jime´nez-Barbero,‡ Javier Can˜ada,‡ and Cristina Vicent*,†
Contribution from the Instituto de Qu´ımica Orga´nica, CSIC, c/ Juan de la CierVa 3,
Madrid 28006, Spain, and Centro de InVestigaciones Biolo´gicas, CSIC, c/ Ramiro de Maeztu 9,
Madrid 28040, Spain
Received February 7, 2005; E-mail: iqocv18@iqog.csic.es
Abstract: Sugar-oligoamides have been designed and synthesized as structurally simple carbohydrate-
based ligands to study carbohydrate-DNA interactions. The general design of the ligands 1-3 has been
done as to favor the bound conformation of Distamycin-type γ-linked covalent dimers which is a hairpin
conformation. Indeed, NMR analysis of the sugar-oligoamides in the free state has indicated the presence
of a percentage of a hairpin conformation in aqueous solution. The DNA binding activity of compounds
1-3 was confirmed by calf thymus DNA (ct-DNA) NMR titration. Interestingly, the binding of the different
sugar-oligoamides seems to be modulated by the sugar configuration. Semiquantitative structural
information about the DNA ligand complexes has been derived from NMR data. A competition experiment
with Netropsin suggested that the sugar-oligoamide 3 bind to DNA in the minor groove. The NMR titrations
of 1-3 with poly(dA-dT) and poly(dG-dC) suggested preferential binding to the ATAT sequence. TR-NOE
NMR experiments for the sugar-oligoamide 3-ct-DNA complex both in D2O and H2O have confirmed the
complex formation and given information on the conformation of the ligand in the bound state. The data
confirmed that the sugar-oligoamide ligand is a hairpin in the bound state. Even more relevant to our
goal, structural information on the conformation around the N-glycosidic linkage has been accessed. Thus,
the sugar asymmetric centers pointing to the NH-amide and N-methyl rims of the molecule have been
characterized.
Introduction
of the carbohydrate became even more relevant with the finding
that Calicheamicin methyl glycoside inhibits the formation of
Carbohydrates are present in many antibiotics and anticancer
drugs that bind DNA.1,2 Carbohydrate minor groove DNA
binders are neutral neoglycoconjugates whose sugar residues
are deoxygenated, thus showing a delicate balance between
hydrophilic and hydrophobic domains.3 The three-dimensional
structure of these oligosaccharides is linear and slightly curved
to fit the minor groove of DNA.4 Traditionally, the glycan chains
of these DNA glycoconjugate binders have been viewed as
molecular elements that control the pharmacokinetics of a drug,
such as absorption, distribution, metabolism, and excretion. This
notion changed a few years ago with the finding that the
carbohydrate residues present in the Calicheamicin antibiotics
partially determine the selectivity of the process.5-9 The role
DNA-protein complexes at micromolar concentrations in a
sequence specific manner.10
However, the lack of knowledge on basic aspects of the
interactions at the origin of selectivity and specificity of
carbohydrate-nucleic acid binding makes it difficult to ef-
fectively design new carbohydrate-nucleic acid binders.11-15
We have previously explored the use of carbohydrate
hydrogen bonding cooperativity16-18 as a tool to bind carbo-
hydrates to the H-bonding motifs present in the grooves of DNA.
(9) Biswas, K.; Pal, S.; Carbeck, J. D.; Kahne, D. J. Am. Chem. Soc. 2000,
122, 8413-8420.
(10) Moreover, it rapidly dissociates preformed complexes not only in vitro but
also in vivo, blocking the expression of a reporter gene under the control
of the transcription factor NFAT. (a) Nicolau, K. C.; Tsay, S.-C.; Suzuki,
T.; Joyce, G. F. J. Am. Chem. Soc. 1992, 114, 7555-7557. (b) Bifulco,
G.; Galeone, A.; Gomez-Paloma, L.; Nicolau, K. C.; Chazin, W. J. J. Am.
Chem. Soc. 1996, 118, 8817-8824. (c) Ho, S. N.; Boyer, S. H.; Schreilser,
S. L.; Danishefsky, S. J.; Crabtree, G. R. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 9203-9207. (d) Sissi, C.; Aiyar, J.; Boyer, S.; Depew, K.;
Danishfsky, S.; Crothers, D. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
10643-10648.
† Instituto de Qu´ımica Orga´nica.
‡ Centro de Investigaciones Biolo´gicas.
(1) Krugh, T. R. Curr. Opin. Struct. Biol. 1994, 4, 351-364.
(2) Kahne, D. Chem. Biol. 1995, 2, 7-12.
(3) Kirschning, A.; Bechthold, A. F.-W.; Rohr, J. Top. Curr. Chem. 1997, 184,
1-84.
(4) Walker, S.; Valentine, K. G.; Kahne, D. J. Am. Chem. Soc. 1990, 112,
6428-6429.
(11) Sakurai, K.; Shinkai, S. J. Am. Chem. Soc. 2000, 122, 4520-4521.
(12) Ding, W.-d.; Ellestad, G. A. J. Am. Chem. Soc. 1991, 113, 6617-6620.
(13) Depew, K. M.; Zeman, S. M.; Boyer, S. H.; Denhart, D. J.; Ikemoto, N.;
Danishefsky, S. J.; Crothers, D. M. Angew. Chem., Int. Ed. Engl. 1996,
35, 2797-2801.
(14) Xuereb, H.; Maletic, M.; Gildersleeve, J.; Pelczer, I.; Kahne, D. J. Am.
Chem. Soc 2000, 122, 1883-1890.
(15) Wu, M.; Stoermer, D.; Tullius, T. D.; Townsend, C. A. J. Am. Chem. Soc
2000, 122, 12884-12885.
(5) Drak, J.; Iwasawa, N.; Danishefsky, S.; Crothers, D. M. Proc. Natl. Acad.
Sci. U.S.A. 1991, 88, 7464-7468.
(6) Zein, N.; Poncin, M.; Nilakantan, R.; Ellestad, G. A. Science 1989, 244,
697-699.
(7) Danishefsky, S. J.; Shair, M. D. J. Org. Chem. 1996, 61, 16-44.
(8) Walker, S.; Murnick, J.; Kahne, D. J. Am. Chem. Soc. 1993, 115, 7954-
7961.
9
9518
J. AM. CHEM. SOC. 2005, 127, 9518-9533
10.1021/ja050794n CCC: $30.25 © 2005 American Chemical Society

Carbohydrate-Based DNA Ligands
A R T I C L E S
Scheme 1. Hairpin/Linear Conformers in Equilibrium
Thus, we have characterized the sugar H-bonding patterns that
efficiently bind CG19 base pairs and phosphate20 in non polar
solvents. In this paper, we present a strategy that extends our
work to natural systems, with the aim of obtaining information
about the structural requirements for DNA groove recognition
by sugars in aqueous media. With this purpose in mind, a series
of sugar-oligoamides have been designed and synthesized as
a tool to study carbohydrate-DNA structural requirements
(Scheme 1). The oligoamide fragment, selected as DNA linker,
resembles those of the naturally occurring antibiotics such as
Distamycin A and Netropsin. These oligoamides are known to
bind DNA in the minor groove as antiparallel dimers.21,22
Moreover, it has been shown that covalently linking such dimers
dramatically increases the DNA binding affinity through the
opportunity for face-to-face pairing of the heterocyclic rings.
These investigations have culminated in the development of
ring-pairing rules by which oligoamides can be constructed from
N-methylpyrrole, -imidazole, and -hydroxypyrrole amino acids
with predictable DNA sequence selective binding.23-26 Recently,
these oligoamides have been probed to inhibit transcription in
vivo.27-29 The design of our carbohydrate-containing DNA
ligands is based on these previous data. Thus, we have selected
the side-by-side pairing of pyrrole amino acids (Py) covalently
joined head-to-tail by a γ-amino butyric acid linker (γ). We
expected the Py/Py pair30 to be the minimum oligoamide
structure able to retain DNA binding ability, thus allowing
DNA-carbohydrate proximity. Additionally, the γ linker is
known to be the most suitable alkyl chain length for hairpin
formation.31 We have included an indole ring32 at the N-terminus
in order to drive the linear/hairpin equilibrium (Scheme 1)
toward the folded conformation through carbohydrate/aromatic
CH-π interactions33,34 which is the conformation which
interacts with DNA. In addition, the indole-2-carboxamide was
selected as an extended mimic of the pyrrole amino acid, which
maintains the necessary crescent geometry of the oligoamide.
Clearly, the configuration and conformation of the bond that
links the sugar to the oligoamide will be a decisive factor in
determining the position of the equilibrium. We decided to
prepare the C-1 amido-sugars which, taking into account the
predicted conformation for anomeric amides,35,36 could generate
the hairpin conformer. The C-1 linkage, in principle, would
make possible generating a preferred orientation of the sugar
within the hairpin structure and, hence, “conformational restric-
tion”. The different relative configuration of the asymmetric
centers in the sugar can also play a role in defining the face
selection, and thus, a particular orientation of the sugar will be
exposed toward the floor of the minor groove. This feature will
be relevant in the further design of sugars with particular
hydrophilic or hydrophobic residues oriented toward the minor
groove.
Thus, starting from the conformation corresponding to
Dervan’s oligoamides bound to DNA in solution,37 molecular
models of the sugar-oligoamides with the oligoamide in C-1
were built and minimized. Careful inspection of the models
obtained with the anomeric linkage R and â strongly suggests
that a â-linkage in the sugar moiety is required in order to allow
a proper pyranose-indole stacking (see Figure 1).
As a first step, sugar-oligoamides from galactose â-Gal-Py-
γ-Py-Ind 1, glucose â-Glc-Py-γ-Py-Ind 2, and xylose â-Xyl-
Py-γ-Py-Ind 3, were prepared (see Figure 2). The conformation
of 1-3 in the free state was analyzed by NMR methods. Then,
the DNA binding activity of the sugar-oligoamides has been
tested. Thus, NMR titration experiments with ct-DNA, poly (dA-
dT), and poly (dG-dC) conclusively show that all synthesized
ligands bind to DNA. Moreover, a qualitative description of
the mode of interaction has been accomplished through the use
TR-NOESY NMR experiments.
(16) Lopez de la Paz, M.; Ellis, G.; Penade´s, S.; Vicent, C. Tetrahedron Lett.
1997, 38, 1659-1662.
(17) Lopez de la Paz, M.; Jime´nez-Barbero, J.; Vicent, C. Chem. Commun. 1998,
465-466.
(18) Luque, F. J.; Lo´pez, J. M.; Lo´pez de la Paz, M.; Vicent, C.; Orozco, M. J.
Phys. Chem. A 1998, 102, 6690-6696.
(19) Lo´pez de la Paz, M.; Gonza´lez, C.; Vicent, C. Chem. Commun. 2000, 411-
412.
(20) Mun˜oz, E. M.; Lo´pez de la Paz, M.; Jime´nez-Barbero, J.; Ellis, G.; Pe´rez,
M.; Vicent, C. Chem.sEur. J. 2002, 8, 1908-1914.
(21) Pelton, J. G.; Wemmer, D. E. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 5723-
5727.
(22) Pelton, J. G.; Wemmer, D. E. J. Am. Chem. Soc. 1990, 112, 1393-1399.
(23) Dervan, P. B.; Bu¨rli, R. W. Curr. Opin. Chem. Biol. 1999, 3, 688-693.
(24) White, S.; Szewczyk, J. W.; Turner, J. M.; Baird, E. E.; Dervan, P. B.
Nature 1998, 392, 468-471.
(33) The indole sidechain of the tryptophan amino acid is present in many lectin
binding sites and is known to contribute to sugar binding through CH-
interactions. (a) Asensio, J. L.; Can˜ada, F. J.; Siebert, H.-C.; Laynez, J.;
Poveda, A.; Nieto, P. N.; Soedjanaamadja, U. M.; Gabius, H.-J.; Jimene´z-
Barbero, J. Chem. Biol. 2000, 7, 529-543. (b) Jimene´z-Barbero, J.; Asensio,
J. L.; Can˜ada, F. J.; Poveda, A. Curr. Opin. Struct. Biol. 1999, 9, 549-
555.
(25) Geierstanger, B. H.; Mrksich, M.; Dervan, P. B.; Wemmer, D. E. Science
1994, 266, 646-650.
(26) Wemmer, D. E.; Dervan, P. B. Curr. Opin. Struct. Biol. 1997, 7, 355-
361.
(27) Janssen, S.; Durussel, T.; Laemmli, U. K. Mol. Cell 2000, 6, 999-1011.
(28) Janssen, S.; Cuvier, O.; Mu¨ller, M.; Laemmli, U. K. Mol. Cell 2000, 6,
1013-1024.
(34) Muraki, M. Protein Pept. Lett. 2002, 9, 195-209.
(29) Henikoff, S.; Vermaak, D. Cell 2000, 103, 695-698.
(30) Dervan, P. B. Science 1986, 232, 464-471.
(31) Mrksich, M.; Parks, M. E.; Dervan, P. B. J. Am. Chem. Soc. 1994, 116,
7983-7988.
(35) Avalos, M.; Babiano, R.; Carretero, M. J.; Cintas, P.; Higes, F. J.; Jime´nez,
J. L.; Palacios, J. C. Tetrahedron 1998, 54, 615-628.
(36) Masuda, M.; Shimizu, T. Carbohydr. Res. 2000, 326, 56-66.
(37) Hawkins, C. A.; Pela´ez de Clairac, R.; Dominey, R. N.; Baird, E. E.; White,
S.; Dervan, P. B.; Wemmer, D. E. J. Am. Chem. Soc. 2000, 122, 5235-
5243.
(32) Schnell, J. R.; Ketchem, R. R.; Boger, D. L.; Chazin, W. J. J. Am. Chem.
Soc. 1999, 121, 5645-5652.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9519

A R T I C L E S
Martin et al.
Figure 1. Molecular model of R-Xyl-Py-γ-Py-Ind (a) and â-Xyl-Py-γ-Py-Ind 3 (b). Both views (left and right) are related by a 90° rotation around the
x-axis.
Figure 2. Sugar-oligoamides and oligamides’ models.
Results and Discussion
Thus, commercial galactose pentaacetate was converted to
the C-1 azide38,39 and reduced (H2, Pd) to the corresponding
amine 1240 (96%, two steps). 1-Methyl-4-nitropyrrole-2-car-
boxylic acid 1341 was prepared for coupling with protected
amine 12 by activation of the acid with DIPC and HOBt or by
conversion to the acid chloride 1442,43 to afford the target
nitropyrrole amide 9 (86% and 87%, respectively). Similarly,
the â-galactosylamine 1542,43 and the â-xylosylamine 16 were
converted to the amides 10 and 11, respectively. The nitro
function represents a latent amino group (revealed by hydro-
genation), so avoiding excessive manipulation of the unstable
aminopyrroles. Nitropyrrole 9 (an intermediate of the target
sugar oligoamide 3) also serves as a precursor of the galactose
strand model compound 4 through hydrogenation and N-
acetylation (nitro- to acetamido-) to give 17, before acetate
cleavage to afford model 4. Similarly, glucosyl- and xylosyl-
pyrroles 10 and 11 were converted to the corresponding sugar
strand models 5 and 6 via acetamides 18 and 19.
Accordingly, we have prepared the sugar-oligoamides â-Gal-
Py-γ-Py-Ind 1, â-Glc-Py-γ-Py-Ind 2, and â-Xyl-Py-γ-Py-Ind
3 (Figure 2) by a convergent route. Thus, the chains are built
up from two separate suitably protected strands, forming the
final amide linkage (arrowed) at a late stage. These sugar
oligoamides were selected to determine the influence of the
sugar stereochemistry on the preferred conformation of these
glycoconjugates in the free state and once bound to DNA
through subtle differences in the sugar moieties/configuration.
The galactose â-Gal-Py-γ-Py-Ind 1 presents an axially oriented
OH at position 4 of the sugar â face, while glucose derivative
2 places a hydrogen atom at this position. Xylose oligoamide 3
is a glucose analogue that lacks the 6-hydroxymethyl group,
thus increasing the number of C-H in the â face.
Additionally, the corresponding sugar strand amides â-Gal-
Py-NHAc 4, â-Glc-Py-NHAc 5, and â-Xyl-Py-NHAc 6, the
indole strand acid HO-γ-Py-Ind 7, and the analogous amide
MeNH-γ-Py-Ind 8 were prepared for comparative purposes
(Figure 2). Those fragments common to all of the sugar
oligoamides 1-3 were prepared as models of the unfolded
conformation. Thus, conformational analysis of compounds 1-8
was carried out in D2O/[D6]-acetone (9:1), D2O/H2O mixtures
(phosphate buffer).
Synthesis of the indole strand acid 7 started with ester 20,
prepared from commercial materials, by the haloform methodol-
ogy of Xiao and co-workers44 (Scheme 3). Hydrogenation of
(38) Paulsen, H.; Gyo¨rgydea´k, Z.; Friedman, M. Chem. Ber. 1974, 107, 1568-
1578.
(39) Tropper, F. D.; Andersson, F. O.; Braun, S.; Roy, R. Synthesis 1992, 618-
620.
(40) Sabesan, S. Tetrahedron Lett. 1997, 38, 3127-3130.
(41) Baird, E. E.; Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6141-6146.
(42) Lown, J. W.; Krowicki, K. J. Org. Chem. 1985, 50, 3774-3779.
(43) Mrksich, M.; Dervan, P. B. J. Am. Chem. Soc. 1993, 115, 9892-9899.
Synthesis. Our convergent synthesis started with the prepara-
tion of the sugar strand fragments 9, 10, and 11 (Scheme 2).
9
9520 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
Scheme 2 a
a (i) DIPC, HOBt, 13; (ii) 14, CHCl3, Et3N, ∆; (iii) SOCl2, ∆, 1 h; (iv) H2, Pd-C, MeOH or DCM, then AcCl or Ac2O, Et3N, ∆; (v) NaOMe, MeOH,
rt, 1 min, then Amberlite IR-120 (H+) to pH ) 6.
Scheme 3 a
a (i) H2, 10%, Pd-C; (ii) DIPC, HOBt, DCM or DMF; (iii) aq NaOH (1 M), MeOH, 1 min then aq HCl (to pH ) 2); (iv) DIPC, HOBt, H2NMe‚HCl,
Et3N; (vi) NaOMe, MeOH, 1 min then Amberlite IR-120 (H+) to pH ) 6.
20 gave aminopyrrole 21, used immediately in a carbodiimide-
mediated coupling with indole-2-carboxylic acid 22 to afford
the ester oligoamide 23. Chromatography did not remove all
traces of the rest of coupling byproduct (diisopropyl urea:
DIPU), but hydrolysis of the crude ester 23 (aq NaOH, 1 min)
afforded the acid 7, isolated in pure form. With the required
fragments now in hand, the sugar nitropyrroles 9, 10, or 11 were
separately reduced and the crude amines added to the preformed
activated ester of the acid 7 to form the desired protected
oligoamides [(AcO)4-â-Gal-Py-γ-Py-Ind 1(OAc), (AcO)4-â-Glc-
Py-γ-Py-Ind 2(OAc), and (AcO)4-â-Xyl-Py-γ-Py-Ind 3(OAc)].
Instantaneous acetate cleavage with sodium methoxide afforded
the target â-glycosyl-oligoamides 1, 2, and 3, respectively. The
indole strand acid 7 was also used to prepare our final model
compound 8 by amide coupling with methylamine.
Structural Characterization and Conformational Studies
of 1-8 in Aqueous Solution. It is well-known that the 3D
structure and dynamic behavior of a ligand in the free state play
an important role in the recognition event. Thus, it has been
suggested that, in those cases where the ligand is flexible in
solution, the loss of conformational freedom associated to
binding constitutes one of the main contributions to the entropic
barrier of the process.45 In this sense, a possible strategy in the
design of higher affinity ligands would be to chemically block
the ligand in its bioactive conformation.
Recent studies by Wemmer et al. have proved that DNA
ligands such as Im-Py-Py-γ-Py-Py-Py-â-Dp adopt a hairpin
geometry in the bound state.37,46,47 This hairpin structure was
defined by NOEs between H-3 protons of the rings stacked
(45) Williams, D. H.; Stephens, E.; O’Brien, D. P.; Zhou, M. Angew. Chem.,
Int. Ed. 2004, 43, 6596-6616.
(46) Pilch, D. S.; Poklar, N.; Gelfand, C. A.; Law, S. M.; Breslauer, K. J.; Baird,
E. E.; Dervan, P. B. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8306-8311.
(44) Xiao, J.; Yuan, G.; Huang, W.; Chan, A. S. C.; Lee, K.-L. D. J. Org. Chem.
2000, 65, 5506-5513.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9521

A R T I C L E S
Martin et al.
Table 1. 1H NMR Chemically Induced Shift (CIS) Differences
between Sugar-Oligoamides Resonance of 1-3 and Their
Corresponding Sugar Strand 4-6 and Indol Strand 7 in D2O/
Acetone [9:1]
Gal
∆δH [1
∆δH [(1
Glc
∆δH [2
∆δH [(2
Xyl
∆δH [3
∆δH [(3
−
4]
−
5]
−
6]
−7)]
−7)]
−7)]
Figure 3. Conformation of the N-glycosidic linkage: face selection.
In3
(+0.031)
(-0.047)
(-0.044)
(-0.043)
(-0.002)
(+0.028)
(-0.002)
+0.089
+0.063
(0)
+0.014
(+0.082)
(+0.136)
(+0.217)
(-0.032)
(-0.052)
(-0.041)
(-0.042)
(-0.027)
(-0.043)
(-0.093)
-0.041
-0.071
(-0.072)
-0.091
(+0.060)
(+0.124)
(+0.187)
-0.019
-0.051
-0.050
(-0.036)
(+0.008)
(+0.009)
(+0.006)
(+0.009)
(-0.145)
(-0.150)
-0.113
-0.197
(-0.101)
-0.147
?
(+0.166)
(+0.190)
-0.137
-0.150
-0.158
In4
across the groove, as well as analogous contacts between H-5s
and the N-methyls. Thus, a crescent structure in which the
methyls of Py and Im are in one rim (N-methyl rim) pointing
outside the groove and the NH amides and H-3 of Py and Im
are in the opposite rim (NH-amide rim) pointing toward the
floor of the groove has been described. Interestingly, the free
state conformation of these molecules has never been studied
in solution. In this context, our sugar-oligoamide derivatives
have been carefully designed to fold in a hairpin structure
additionally stabilized by sugar-indole CH-π interactions.
To verify the presence of the bioactive (putative bound state)
hairpin structure for the sugar-oligoamides in the free state, a
conformational study of 1-3 has been carried out by NMR, in
conjunction with molecular mechanics calculations.
In5
In6
In7
P3A
P5A
P3B
P5B
CH3
CH3
CH2
CH2
CH2
A
B
a
b
c
sugar
0.048-0.025
H1
+0.019
Table 2. 1H NMR Chemically Induced Shift (CIS) Differences
between Sugar-Oligoamide Resonances, Excluding the
Carbohydrate Protons of 1-3 and Their Corresponding Sugar
Strands 4-6 and Indol Strand 7 in D2O (Phosphate Buffer)
In addition, we can envisage several different conformations
depending on the relative orientation of the â-sugar moiety and
the indole ring (Figure 3).
Gal
∆δH [1
∆δH [(1
Glc
∆δH [2
∆δH [(2
Xyl
∆δH [3
∆δH [(3
Presumably, the most populated structure in solution will be
dependent on the sugar stereochemistry. The structural analysis
of compounds 1-3 solution was carried out by NMR methods.
−
4]
−
5]
−
6]
−7)]
−7)]
−7)]
In3
(-0.059)
(0)
(0)
(-0.047)
(+0.004)
(+0.008)
(+0.001)
(+0.006)
(-0.238)
(-0.157)
-0.164
-0.264
(-0.100)
-0.214
?
(-0.049)
(+0.008)
(+0.011)
(+0.004)
(+0.009)
(-0.217)
(-0.170)
-0.166
-0.285
(-0.110)
-0.199
(+0.099)
(+0.193)
(+0.170)
1
As a first step, a full assignment of the H NMR spectra of
In4
In5
1-8 in D2O/[D6]-acetone (9:1), D2O and H2O was obtained
using two-dimensional experiments (HMQC, TOCSY, and
NOESY).
In6
(-0.006)
(+0.001)
(-0.270)
(-0.142)
-0.149
-0.262
(-0.092)
-0.214
?
In7
P3A
P5A
P3B
P5B
CH3
CH3
CH2
CH2
CH2
1H NMR resonances spectra in D2O/[D6]-acetone and
H2O/[D6]-acetone (phosphate buffer, 10 mM, pH ) 7) for NH
assignment were registered for a total assignment of com-
pounds 1-8 (see Tables A-D in the Supporting Information).
The first evidence of the existence of a percentage of a particular
hairpin conformation in the free state was obtained by com-
parison of the chemical shifts of the sugar-oligoamides 1-3
with those of the indole strand amide 8 (which is common to
the three sugar-oligoamides) and the sugar strand models 4-6.
It is well-known that the proximity of aromatic rings induces
A
B
a
b
(+0.194)
(+0.167)
(+0.183)
(+0.170)
c
strand models and showed the largest shifts within all the proton
resonances. The same behavior was found for the methyl N-Py
resonances.
1
chemical shift changes in the H NMR spectrum. Tables 1-4
show the ∆δH of the proton resonances of each of the sugar
oligoamides 1-3 and the corresponding sugar strand amides
4-6 and the indole strand acid 7 (the indole strand amide 8
was not used due to the low solubility of the product). The
chemically induced shifts (CISs) for all the resonances follow
the same trend in both D2O [D6]-acetone (Table 1) and D2O
solution (Table 2), but the relative differences are much larger
in water. Proton resonances of the Py rings (P3A, P3B, P5A, and
P5B) were shielded in comparison with their correspondent single
The sign of the chemically induced shift (CIS) of the Py
signals (P3, P5, and CH3) is compatible with the presence of
some extent of an intramolecular folded conformation in which
the aromatic residues of each strand are stacked. The comparison
of ∆δ of the three sugar-oligoamides did not show significant
differences depending on the sugar stereochemistry derivative
at the C-terminus (Table 2).
Of even greater relevance to the design were the CISs of the
sugar ring resonances when the sugar-oligoamides (1-3) and
the correspondent sugar strand derivatives (4-6) were compared
(Table 3) in D2O.
(47) Pelaez Lamanie de Clairac, R. P. L.; Geierstanger, B. H.; Mrksich, M.;
Dervan, P. B.; Wemmer, D. E. J. Am. Chem. Soc. 1997, 119, 7909-7916.
9
9522 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
Table 3. 1H NMR Chemically Induced Shift (CIS) Differences
between Carbohydrate Residue of Sugar-Oligoamides 1-3
Resonances and Their Corresponding Sugar Strand 4-6 in D2O
absence of an NOE between NH5 and H-1 supports the presence
of a major conformation in solution.
In principle, the above-described features in the NMR spectra
of 1-3 could be additionally due to intermolecular oligomer-
ization, since chemically induced shifts could also be expected
for either a head-to-tail dimer or an oligomer of the sugar-
amides. We ruled out this possibility, through dilution experi-
ments (in D2O/[D6]-acetone and D2O), which showed that the
δ of 1-3 did not change in the concentration range from 10
mM to 0.1 mM, where the self-associated oligomers should
become less favorable. This confirms that the chemically
induced shifts found in the proton resonances of 1-3 are due
to intramolecular folding. Dilution experiments were also
performed in the model compounds 4-7 to compare with their
corresponding sugar-oligoamides, and with the indole strand
acid 7. Only the acid 7 shows evidence of self-association in
D2O, with a small dimerization constant (Ka ) 100 M-1).
The next step was to characterize the geometry of the folded
conformation present in solution. With this purpose, NOESY
and ROESY experiments were carried out with 1-3 in D2O/
[D6]-acetone, D2O, and H2O/[D6]-acetone solutions. The small
intensity of the measured NOEs obtained in D2O, as well as
the strong overlapping of the sugar resonances (mainly in â-Gal-
Py-γ-Py-Ind 1 and â-Xyl-Py-γ-Py-Ind 3), made the structural
analysis difficult.
Gal
Glc
Xyl
∆δH [1−4]
∆δH [2−5]
∆δH [3−6]
H-1
H-2
H-3
H-4
H-5
H-5′
H-6
H-6′
-0.33
-0.07
a
-0.06
a
-0.48
-0.30
-0.13
-0.12
-0.06
-0.31
-0.14
-0.10
-0.31
-0.26
-0.29
-0.10
0
a Overlap.
Table 4. 1H NMR Chemically Induced Shift (CIS) Differences
between Sugar-Oligoamide Resonances of 1-3 and Their
Corresponding Sugar Strands 4-6 and Indol Strand 7 in H2O/
[D6]-Acetone [9:1] for NH Resonances
Gal
∆δH [1
∆δH [(1
Glc
∆δH [2
∆δH [(2
Xyl
∆δH [3
∆δH [(3
−
4]
−
5]
−
6]
−7)]
−7)]
−7)]
NH1
NH2
NH3
NH4
NH5
(-0.162)
(-0.210)
(-0.246)
0.008
(-0.149)
(-0.196)
(-0.241)
-0.010
(-0.107)
(-0.154)
(-0.210)
0.066
-0.205
-0.106
-0.201
All of the sugar pyranose resonances were shielded in the
sugar-oligoamides, reflecting their spatial proximity to the
indole ring. The anomeric proton resonances present signifi-
cantly different behaviors in both solvents, depending on the
sugar residue. In D2O/[D6]-acetone (Table 1) H-1 of the
galactose-oligoamide 1 compared with the sugar strand model
4 is +0.019 ppm deshielded. In contrast, the H-1 protons of
oligoamides 2 and 3 are shielded. In D2O (Table 3), the H-1
proton of 1-3 is shielded to a similar extent. This suggests either
a different relative orientation of the sugar residue with respect
to the indole ring or a different degree of hairpin conformation,
depending on the solvent. The largest shifts were found for
proton resonances of the pyranose rings of Xyl 3, while the
smaller for these particular resonances were observed for Gal
1.
In addition to the γ-linker, several potentially flexible bonds
have to be taken in account for structural analysis of 1-3. The
conformational preferences around these bonds (Figure 4a)
would determine the precise nature of the hairpin structures
within the strands. Nevertheless, the experiments in H2O/[D6]-
acetone allowed tracing the connectivity shown in Figure 4b.
Thus, they permitted to unambiguously establish the most
populated orientation for these linkages in solution.
These intrastrand NOEs allowed confirming the crescent
conformation of the three sugar-oligoamides (1-3). Similar
intrastrand NOEs were described for γ-linked oligoamide
covalent dimers in the bound state conformation.37,46,47
Particularly interesting is the presence of an NOE between
NH2 and In3. Thus, the indole In3 proton is pointing toward the
NH-amide rim, as suggested by the molecular models. In this
sense, the indole residue can be considered as an extended mimic
of Py.
Table 4 shows the ∆δ of the NH resonances of 1-3 in H2O/
[D6]-acetone, in comparison with the corresponding model
compounds 4-7.
Finally, NOESY experiments in D2O/[D6]-acetone allowed
the determination of the 3D structure of the designed oligo-
amides in solution.
All NH resonances of 1-3 were shielded with respect to the
model except NH4, which shows a very small CIS. In fact, this
proton follows a different trend depending on the sugar. Again,
these upfield shifts (from -0.106 to -0.246 ppm) are consistent
with a folded conformation, in which the aromatic residues of
each strand are facing each other, thus affecting the NH
resonances. The coupling constant values found in NH5
resonances for 1-3, and the model compounds 4-6, which
correspond to the proton amide in the sugar anomeric position
(9.2-8.5 Hz) (Table D in Supporting Information) are consistent
with a trans (180°) or syn (0°) orientation of NH5 respect to
H-1 of the sugar pyranose ring allowing an almost parallel
orientation between the plane defined by the sugar pyranose
and the indole ring.35 This is in accordance with the predicted
geometry (Figure 2) for the â-anomer sugar-oligoamide hairpin.
Based on Avalos et al.35 there is a preferred anti conformation
for structurally similar amido sugars. In our case, the presence
of an NOE between NH5 and H-2 of the pyranose ring and the
These experiments showed interstrand NOEs (Table E in the
Supporting Information) characteristics of the hairpin conforma-
tion, described by Wemmer and co-workers in the bound state
for γ-linked oligoamides.37,46,47
Thus, spatial proximity among P3A-P3B, P5A-CH3 , and
B
P5B-CH3 was found in solution in the free state for 1-3
A
(Figure 5), suggesting that the bioactive hairpin structure,
previously described for the structurally more complicated
minor-groove-DNA-binding Distamycin-type covalent di-
mers,37,46,47 is not significantly disturbed by the presence of the
sugar moiety. It is remarkable that, in addition to these contacts,
unambiguous NOEs between the sugar and the indole ring could
be assigned. Thus, NOEs between the indole proton resonances
and the sugar pyranose proton resonances present in the â-face
of Glc in 2 and Xyl in 3 (Table E in Supporting Information)
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9523

A R T I C L E S
Martin et al.
Figure 4. (a) Possible bond rotations of sugar-oligoamide. (b) Intrastrand NOEs found in the NOESY spectra.
Figure 5. Interstrand NOEs for oligoamides 2 and 3.
Table 5. 1H NMR Titration of Compounds 1-3, 6, and 7 with Calf Thymus DNAa
â
-Gal-Py-
γ
-Py-In 1
â
-Glc-Py-
γ
-Py-In 2
â
-Xyl-Py-
γ
-Py-In 3
â
-Xyl-Py-NHAc 6
In
−
Py-
γ
-OH 7
Netropsin
mM
mM
mM
mM
mM
mM
concentration
of ct-DNA
2.6
2
0.36
>3.5
>3.5
0.03
a Concentration of the solution. Conditions: D2O, θ ) 26 °C, [oligo] )10-4 M (0.5 mL), [DNA] )2 mg/0.7 mL
were detected. More specifically, the â-face proton resonances
H-2, H-4, and H-6a/H-6b of Glc 2 sugar-oligoamide show clear
contacts with In3, In4, In5, and In7. In a similar way, NOEs
between sugar H-2 and H-5eq and In4, In5, and In7 were found
for the Xyl 3 derivative. Two different conclusions can be drawn
from these results. First, they constitute unequivocal evidence
of the presence of a percentage of folded bioactive conformation
in aqueous solution. Additionally, it allowed unambiguously
defining the precise geometry of the sugar-indole interaction:
the aromatic ring mainly stacks against the â-face of the sugar
residue.48 This is similar to that found in the bound state for
Dystamicin-type covalent dimer minor groove binders.
Indeed, the conformational analysis of sugar-oligoamides
1-3 in the free state shows that the water soluble ligands present
a percentage of hairpin conformation in solution in which an
NH-amide rim and an N-methyl rim has been characterized
(Figure 6). Regarding the sugar residue, the â-face of the
pyranose ring is close to the indole ring in the hairpin
conformation. Thus, C-2 and C-3 of the pyranose are pointing
to the NH-amide rim while C-4, C-5, and O-5 in xylose and
C-4, C-5, O-5, and C-6 in glucose are facing the N-methyl rim
of the sugar-oligoamide.
Increasing amounts of ct-DNA were added to a constant con-
centration solution of the oligoamide (1-3). In principle, de-
pending on the kinetics of the free/bound exchange process,
DNA binding could affect the chemical shifts, line intensities,
and/or the line widths of the ligand resonances. These perturba-
tions can be used to monitor the binding process. Indeed, a sig-
nificant line broadening of resonances from compounds 1-3,
which eventually disappeared below the noise level, is observed
upon addition of DNA. Table 5 shows the amount of ct-DNA
(in base-pair, bp) needed to make disappear from the spectra at
least 95% of the resonance signals of the ligand (see Scheme 4).
The sugar-oligoamides 1-3 and the model compounds Xyl-
Py-NHAc 6 and HO-γ-Py-Ind 7 were used in different binding
experiments.
Interestingly, there are significant differences in the amount
of ct-DNA needed to produce a similar effect on the ligand
signals, depending on the nature of the sugar moiety. Indeed,
the Xyl-oligoamide 3 proton resonances disappeared from the
NMR spectrum upon addition of ct-DNA (0.36 mM bp). In
contrast, 2 and 1 reached the same situation with much higher
concentrations (2 mM and 2.6 mM in bp, respectively). Thus,
there is nearly an order of magnitude difference in the required
DNA concentration between â-D-Xyl-Py-γ-Py-Ind 3 and â-D-
Gal-Py-γ-Py-Ind 1 for the same effect. Although only qualita-
tive, this observation strongly suggests that DNA binding affinity
and/or specificity is rather different for ligands 1, 2, and 3. Thus,
DNA Interaction. As a further step, we have explored the
binding of sugar-oligoamides 1-3 to calf-thymus DNA (ct-
1
DNA). First, a H NMR titration experiment was carried out.
(48) Bernardi, A.; Arosio, D.; Potenza, D.; Sanchez-Medina, I.; Mari, S.; Canada,
F. J.; Jimenez-Barbero, J. Chem.sEur. J. 2004, 10, 4395-4406.
9
9524 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
Scheme 4. Titration of 3 with ct-DNA.
subtle changes on the sugar residue structure may have
implications on DNA binding.
yl-R-D-glucopyranoside were present in the same concentration
(see Supporting Information). Upon addition of ct-DNA at a
constant concentration of both substrates, while â-Glc-Py-γ-
Py-Ind 2 resonances start to broaden and disappear, the
O-methyl-R-D-glucopyranoside resonances remained unaffected.
As expected, the O-methyl-R-D-glucopyranoside is not a DNA
substrate. This is probing that the above-mentioned effect
observed on 1-3 resonance signals upon addition of ct-DNA
is directly related to DNA binding.
There are also differences in the behavior of the individual
proton resonances of the sugar-oligoamides upon addition of
ct-DNA. Thus, proton resonances from the indole moiety and
pyrrol (P3A and P5A) are earlier affected (they broaden) upon
addition of ct-DNA than the rest of them. In contrast, the CH3
N-pyrrol resonances only disappear, below the noise level, after
the latest additions of ct-DNA and, again, CH3A does it earlier
than CH3B (see Supporting Information). This general behavior
of the three sugar-oligoamides (1-3) is suggesting that the
NH-amide rim in the bound state is closer to the ct-DNA than
the N-methyl rim and that the indole strand is closer to the ct-
DNA than the sugar strand in the bound state (Figure 6). This
fact seems to be in agreement with a bound state conformation
and binding mode similar to those described for γ-linked
oligoamide covalent dimers.37,46,47
This control experiment was also carried out in an indepen-
dent manner, by addition of ct-DNA to a solution of O-methyl-
R-D-glucopyranoside, in the same experimental conditions
described for 1-3. Even at high concentrations of DNA (more
than 4mM in bp) the O-methyl-R-D-glucopyranoside proton
resonances did not disappear from the spectra.
Interestingly, under the conditions just mentioned for the
sugar-oligoamide titrations, neither the sugar-oligoamide
models (4-6) nor the indole strand acid Ind-Py-γ-COOH (7)
are DNA binders (after addition of more than 4 mM ct-DNA
As a control experiment, two ct-DNA titrations were also
carried out. In both of them, O-methyl-R-D-glucopyranoside, a
simple sugar, lacking of the DNA-binding oligoamide linker,
was used. The first experiment was a titration experiment in
which both sugar-oligoamide â-Glc-Py-γ-Py-Ind 2 and O-meth-
1
solution the resonances did not disappear from the H NMR
spectra).
All these evidences are indicating that these isolated fragments
(indol strand and the sugar strand) do not show enough contact
sites for efficient binding.
As a positive control experiment, the DNA binding experi-
ment was performed using Netropsin as ligand, a well-known
effective minor groove binder. It binds ct-DNA with a Ka )
428 × 105 M-1 49
.
Thus, a 0.1 mM solution of Netropsin was titrated with a
Figure 6. Schematic representation of â-Glc-Py-γ-Py-Ind 2 in solution.
solution of ct-DNA which was 30 times more diluted than those
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9525

A R T I C L E S
Martin et al.
Figure 7. Competition experiment between â-Xyl-Py-γ-Py-Ind 3 and Netropsin; (A) 1H NMR spectra of 3 (0.1mM), (B) 3 in the presence of ct-DNA
0.23mM bp, (C) 3 in the presence of ct-DNA 0.44mM bp (3-ct-DNA complex), (D) 3-ct-DNA complex upon addition of Netropsin 0.91mM, (E) 3-DNA
complex upon addition of Netropsin 1.27mM, (F) 3 displaced by Netropsin (2 mM).
used for the sugar-oligoamides. Under that condition, 95% of
the proton resonances signals from Netropsin disappear from
the spectra upon ct-DNA addition (0.03 mM). Thus, there is 1
order of magnitude of difference in the required concentration
of DNA needed for the same effect between Netropsin and â-D-
Xyl-Py-γ-Py-Ind 3. This is consistent with an expected more
efficient binding of the positively charged ligand Netropsin,
compared with our neutral sugar-oligoamides 1-3.50 Here
again, there are also differences in the behavior of the individual
proton resonances of the Netropsin upon addition of DNA, and
the behavior is similar to that described above for the sugar-
oligoamides. Thus, proton resonances from the pyrrol moiety
(P3 and P5) are affected earlier than the rest of them. In contrast,
the CH3 N-pyrrol resonances only disappear, below the noise
level, after the latest additions of DNA.
Additionally, an NMR competition experiment was also
performed in order to deduce if Netropsin and 3 were sharing
common binding sites when interacting with ct-DNA. Figure 7
shows the competition experiment following the aromatic
resonances of the substrate (between 6.0 and 8.0 ppm).
A 0.1 mM solution of 3 was titrated with ct-DNA solution
(2 mg/0.7 mL ) 3.58 mM en bp) until the resonances from 3
disappeared from the NMR spectra (50 µL, 0.44 mM bp),
reflecting the 3-ct-DNA complex formation. Then, a solution
of Netropsin (10 mM) was added stepwise to the (3-ct-DNA)
complex. Upon addition of Netropsin (0.91mM and 1.27 mM,
see Figure 7), the resonances from the sugar-oligoamide 3
started to appear again in the spectra, thus reflecting the
displacement of 3 by Netropsin from the ct-DNA binding site.
This experiment shows that the sugar-oligoamide 3 is a
minor groove DNA binder that is displaced by Netropsin,
suggesting that could share similar binding sites.
The binding experiments of sugar-oligoamides 1 and 3 were
also performed with poly(dA-dT) and poly(dG-dC) polynucleo-
tides to obtain information on the sequence selectivity of
binding. Additionally, Netropsin was also used as a reference
for a structurally related selective minor groove binder. The
concentrations (in base pair) of polynucleotide needed to make
1
95% of the substrate resonances disappear from the H NMR
spectra are shown in Table F in the Supporting Information.
The comparison of titrations of both sugar-oligoamides 1
and 3 with poly(dA-dT) and poly(dG-dC) shows that the amount
of poly(dG-dC) needed to achieve the same effect in the NMR
spectra is 4 (for 1) and 9 (for 3) times the concentration of poly-
(dA-dT), respectively. This is suggesting a more efficient
binding with poly(dA-dT). The same tendency (10 times) is also
found for Netropsin-poly(dA-dT) and poly(dG-dC) titrations
but with smaller amounts of macromolecule. In this last case,
it is in agreement with the known fact that Netropsin binds
preferentially poly(dA-dT) than poly(dG-dC).51 These results
suggest that the positively charged minor groove binder,
Netropsin binds an ATAT sequence more effectively than our
neutral sugar-oligoamides (1 and 3). Nevertheless, our sugar-
(51) Ka values of Netropsin-poly(dA-dT) ) 22 000 × 105 M50,52 and Ka
Netropsin-poly(dG-dC) ) 1.61 × 105 M.52 It is important to remark that
the experimental conditions in which Netropsin-poly(dG-dC) and Netrop-
sin-poly(dA-dT) Ka values are reported are not comparable with our
experimental conditions. The lack of salt in our experiments is expected to
enhance electrostatic nonspecific interactions. In fact, some extent of
nonspecific binding cannot be excluded under these conditions and would
explain the small amount of macromolecule concentration needed to achieve
binding.
(49) Chaires, J. B. Biopolymers 1997, 44, 201-215.
(50) The reported Ka values are described in experimental conditions not
comparable with this NMR titration.
9
9526 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
Figure 8. Interstrand and intrastrand NOEs for sugar-oligoamide 3 in the bound state in H2O/D2O (15%).
oligoamides (1 and 3) present also some preference toward the
poly(dA-dT) sequence for binding.
Regarding the comparison between both sugar-oligoamides,
â-Gal-Py-γ-Py-Ind 1 needs twice the concentration of poly-
(dA-dT) than 3. These results show that ligands â-Gal-Py-γ-
Py-Ind 1 and â-Xyl-Py-γ-Py-Ind 3 present a small, but
detectable, distinct behavior in binding toward poly(dA-dT)
which should be attributed to the differences that the carbohy-
drate structure (â-Gal and â-Xyl) confers to the general structure
of the hairpin ligand.
Finally, TR-NOESY experiments were carried out under dif-
ferent experimental conditions (temperature, ligand/DNA ratio),
to analyze the bound state sugar-oligoamide structures in the
complexes with ct-DNA. The TR-NOESY is a regular NOESY
experiment, but it is applied to a big receptor/ligand system in
dynamic exchange, in which the ligand is present in excess. In
Figure 9. Schematic representation of â-Glc-Py-γ-Py-Ind 2 in the bound
general, for bound ligands, which exchange with the free state
at fast rate, exchange transferred (TR-NOE) experiments provide
an adequate means to determine the conformation of the bound
ligand. In complexes involving large molecules, cross relaxation
rates of the bound compound (σB) (which depend on the inter-
proton distances of the bound state, the spectrometer frequency,
and on the correlation time of the bound ligand) are opposite
in sign to those of the free one (σF) and produce negative NOEs.
Therefore, the existence of binding may be easily deduced from
visual inspection, since NOEs for small molecules are positive
or close to zero (as 1-3). Because of the relatively low values
of cross-relaxation and longitudinal rates in diamagnetic systems
(<10 s-1), the experiment is applicable under certain kinetic
conditions (k-1 . σB, where σB is the cross relaxation rate for
the bound and k-1 is the off-rate constant).
state.
amide resonances and the next amino acid residue resonances
(intrastrand) confirmed that the conformation in the bound state
presents an N-H rim and CH3 rim, as detected for the free
ligand (see Figure 8). In this experiment, pyrrol-pyrrol inter-
strand NOEs (see Figure 8) were also detected. Additionally,
NOEs between the sugar â-face proton resonances and the indol
ring (NOEs H-5eq-In6 and H-5eq-In4) were detected, which
suggests that the sugar-oligoamide hairpin bound state con-
formation also places the sugar â-face close to the indol ring.
Thus, regarding the sugar unit, the C-2 and C-3 positions are
directed to the NH-amide rim, while C-4, O-5, C-5, and C-6
are pointing toward the N-methyl rim of the sugar-oligoamide
(see Figure 9).
Thus, the TR-NOESY experiment in D2O of compound 3 in
the presence of ct-DNA was carried out. Negative and intense
NOEs were observed, in contrast with the weak positive NOEs
which 3 presents in the free state. These TR-NOESY data
provide valuable information about the ligand conformation in
the bound state. As previously mentioned, the most relevant
NOEs, in structural terms, are those that define the hairpin
conformation (interstrand NOEs). Thus, clear interstrand con-
tacts between pyrrole A and pyrrole B resonances (NOEs: P3A-
Conclusions
Sugar-oligoamides have been designed and synthesized. The
major aim of this work is to be able to position a structurally
simple monosaccharide at close distance of the DNA minor
groove. Thus, the smallest oligoamide fragment resembling the
known Distamicin-type γ-linked oligoamide dimers has been
selected. It was expected that this design would help the hairpin
conformation (bound state conformation) to be locked in the
free state by π-π interactions and CH-π interactions between
the indol moiety and the sugar pyranose ring.
Indeed, the conformational analysis of sugar-oligoamides
1-3 in the free state shows that these water soluble carbohydrate
ligands present a certain percentage of hairpin conformation in
solution, in which an N-H amide rim and an N-methyl rim
have been characterized. The sugar places its â-face close to
the indole ring in the adopted hairpin conformation. Thus, C-4,
B
P3B, P5A-CH3 , P5B-CH3A) were observed for the sugar-
oligoamide 3 (see Supporting Information). This fact conclusively
proves that compound 3 is bound to ct-DNA and adopts a folded
conformation in the bound state, similar to that observed for
the free state.
The TR-NOESY in H2O also shows NOEs in which the cross-
peak signs are negative, characteristic of binding (see Supporting
Information). Even more relevant, the NOEs between the N-H
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9527

A R T I C L E S
Martin et al.
copyranoside was purchased from Aldrich (Lot 17602-055). H2O for
NMR studies was freshly filtered milli-q water.
O-5, C-5, and C-6 of the pyranose are in the N-methyl rim,
while C-2 and C-3 of the pyranose are pointing toward the N-H
amide rim of the sugar-oligoamide.
Synthesis Section. â-Gal-Py-γ-Py-Ind (1). A solution of (AcO)4-
â-Gal-Py-γ-Py-Ind 1(OAC) (92.0 mg, 0.112 mmol) in MeOH (5 mL)
was treated with a solution of sodium (100 mg, 4.34 mmol) in MeOH
(10 mL) to produce an immediate deeper yellow color, indicative of
reaction to completion. The solution was acidified to pH ) 6 with
Amberlite IR-120 ion-exchange resin (strongly acidic), and the resin
was removed by filtration. Evaporation of the solvent under reduced
pressure afforded the title compound as an off-white solid (40.0 mg,
53%). This hygroscopic residue was dissolved in MeOH (5 mL) and
treated with enough Et2O to cause a white precipitate to form, which
was removed by filtration and dried under vacuum at 40 °C to produce
an analytical sample. Mp ) 200 °C; IR (KBr): ν ) 3600-2500, 1643,
1584, 1530 cm-1; [R]D22 ) +23.7 (c ) 1.35 in [D6] DMSO). 1H NMR
(400 MHz, [D6] DMSO): δ ) 1.79 (m, 2H), 2.28 (m, 2H), 3.22 (m,
2H), 3.30-3.54 (m, 3H), 3.57-3.64 (m, 2H), 3.70 (d, J ) 2.7 Hz,
1H), 3,79 (s, 3H), 3.84 (s, 3H), 4.83 (t, J ) 8.9 Hz, 1H), 6.85 (d, J )
1.7 Hz, 1H), 6.89 (d, J ) 1.7 Hz, 1H), 7.05 (t, J ) 8.0 Hz, 1H), 7.20
(m, 2H), 7.28 (s, 2H), 7.53 (d, J ) 8.2 Hz, 1H), 7.64 (d, J ) 8.2 Hz,
1H), 8.12 (t, J ) 5.7 Hz, 1H), 8.36 (d, J ) 8.8 Hz, 1H), 9.83 (s, 1H),
10.30 (s, 1H), 11.61 (d, J ) 1.5 Hz, 1H). 13C NMR (50 MHz, [D6]
DMSO): δ ) 25.6, 33.3, 36.0, 36.1, 38.2, 60.4, 66.3, 69.1, 74.3, 76.6,
80.1, 102.8, 104.1, 104.5, 112.2, 118.0, 118.4, 119.6, 121.4, 121.5,
121.8, 122.2, 123.2, 127.0, 131.5, 136.5, 141.3, 158.1, 161.1, 161.1,
169.2. MS (ES+): m/z (%): 1304 (6) [2M + H]+, 675 (43) [M +
Na]+, 653 (100) [M + H]+. Elemental analysis calcd (%) for
C31H37N7O9‚2H2O (687.7): C, 54.14; H, 6.01; N, 14.26. Found: C,
54.45; H, 6.02; N, 14.35.
The ability of the ligands to bind ct-DNA has been shown
qualitatively by 1H NMR titration, following the disappearance
of the ligand resonance signals upon increasing addition of ct-
DNA. This method has allowed us to show that 1-3 are ct-
DNA binders while Ind-Py-γ- COOH 7 or sugar-Py-NHAc 4-6
are not ligands in the same conditions. The well-known
structurally related Netropsin minor groove binder has also been
tested by this method, as a control experiment.
The results from the binding experiments suggest, although
only qualitatively, that DNA binding affinity and/or specificity
is different for the three ligands. Thus, subtle changes on the
sugar residue structure may have implications on DNA binding.
â-Xyl-Py-γ-Py-Ind 3 seems to be the best sugar-oligoamide
ct-DNA binder.
Additionally, a competition experiment followed by 1H NMR
between our sugar-oligoamide from xylose 3 and Netropsin
has shown that Netropsin is able to displace 3 from its complex
with ct-DNA, suggesting that 3 is a minor groove binder.
We have also used the 1H NMR titrations to explore sequence
selectivity of binding for sugar-oligoamides 1 and 3 toward
poly (dA-dT) and poly (dG-dC). Comparisons have also been
made with Netropsin in the same conditions. Interestingly, both
Gal 1 and Xyl 3 sugar-oligoamide seem to present a preference
of binding toward the ATAT sequence rather than to GCGC.
(AcO)4-â-Gal-Py-γ-Py-Ind [1(OAc)]. A solution of (AcO)4-â-Gal-
Py-NO2 9 (200 mg, 0.400 mmol) in MeOH (5 mL) and EtOAc (5 mL)
was treated with 10% Pd-C (40 mg, 9 mol %) under H2, 45 psi for 24
h. Once the reaction was finished, the catalyst was removed by filtration
and the solvent evaporated under reduced pressure to afford a pale
brown glassy film which corresponds to the amino derivative of 9.
Separately, a solution of DIPC (63.0 µL, 0.400 mmol), HOBt (54.0
mg, 0.400 mmol), and HO-γ-Py-Ind 7 (147 mg, 0.400 mmol) in DMF
(10 mL) was stirred under argon at room temperature during 24 h before
addition of the solution of the amino derivative of 9 in DMF (5 mL).
After 24 h under these conditions, the solvent was evaporated and the
residue was purified by repeated column chromatography (SiO2, EtOAc,
then SiO2, 10% MeOH in CH2Cl2) to afford an analytical sample of
(AcO)4-Gal-Py-γ-Py-Ind 1(OAc) as an amorphous white solid (140
The most relevant result is that TR-NOESY experiments in
D2O and H2O of xylose sugar-oligoamide 3 in the presence of
ct-DNA have confirmed that the designed carbohydrate ligands
bind ct-DNA (a change in the sign of the NOEs from free to
bound) and the bound conformation presents the same intra-
and interstrand NOEs as those observed in the free state. This
is suggesting that the sugar-oligoamides bind ct-DNA in a
hairpin conformation similar to that found in the free state. More
relevant, the carbohydrate residue places the pyranose ring C-4,
C-5, O-5 and C-6 directed to the N-methyl rim while C-2 and
C-3 of the pyranose are pointing toward the NH-amide rim of
the sugar-oligoamide (see Figure 9). This structural information
of the bound state conformation will be very relevant to the
future design of a sugar residue with particular cooperative HB
motives directed to the minor groove H-bonding centers.19,20,52
mg, 43%). Mp ) 165-167 °C. IR (KBr): ν ) 1748, 1646, 1531, 1248,
1
1232 cm-1. [R]D ) +0.926 (c ) 3.09 in acetone). H NMR (500
23
MHz, [D6] DMSO): δ ) 1.80 (t, J ) 7.5 Hz, 2H), 1.91 (s, 3H), 1.92
(s, 3H), 1.99 (s, 3H), 2.12 (s, 3H), 2.29 (t, J ) 7.5 Hz, 2H), 3.23 (m,
J ) 6.0 Hz, 2H), 3.78 (s, 3H), 3.84 (s, 3H), 4.02 (m, 2H), 4.32 (t, J )
6.0 Hz, 1H), 5.22 (t, J ) 9.3 Hz, 1H), 5.30 (m, 2H), 5.45 (t, J ) 9.3
Hz, 1H), 6.78 (s, 1H), 6.90 (s, 1H), 7.05 (t, J ) 7.5 Hz, 1H), 7.20 (t,
J ) 7.8 Hz, 1H), 7.24 (s, 1H), 7.28 (s, 2H), 7.46 (d, J ) 8.0 Hz, 1H),
7.65 (d, J ) 8.0 Hz, 1H), 8.07 (m, 1H), 8.74 (d, J ) 9.5 Hz, 1H), 9.82
(s, 1H), 10.30 (s, 1H), 11.58 (s, 1H). 13C NMR (125 MHz, [D6]
DMSO): δ ) 20.3, 20.4, 20.4, 20.4, 25.6, 33.2, 36.0, 36.1, 38.2, 61.4,
67.6, 68.3, 71.1, 71.3, 77.5, 102.8, 104.1, 104.9, 112.2, 118.0, 119.2,
119.7, 121.5, 121.5, 121.6, 122.1, 123.3, 123.3, 127.1, 131.6, 136.5,
158.2, 160.9, 161.2, 169.0, 169.3, 169.4, 169.8, 169.9. MS (ES+) m/z
(%): 1663 (8) [2M + Na]+, 1641 (4) [2M + H]+, 843 (89) [M +
Na]+, 821 (100) [M + H]+. Elemental analysis calcd (%) for
C39H45N7O13 (819.8): C, 57.14; H, 5.53; N, 11.96. Found: C, 56.78;
H, 5.60; N, 11.69.
Experimental Section
General. Solvents were purified according to the standard proce-
dures. Melting points are reported in degrees Celcius (uncorrected) with
a microscope Reicher Jung Thermovar. NMR measurements were
recorded on a Varian Gemini-200 (200 MHz), Varian Inova-300 (300
MHz), Varian Inova-400 (400 MHz), Varian Unity-500 (500 MHz),
and Bruker AVANCE 500 MHz spectrometers. Mass spectrometry
analyses were performed with an HP series 1100 MSD. Optical rotations
were measured on a Perkin-Elmer 241MC polarimeter in a 1 dm cell.
Elemental analyses were measured on a Carla Erba CHNS-O EA1108
chromatography system. Thin-layer chromatography was performed on
silica gel on aluminum sheets 60F254 (Merck). Flash chromatography
was performed on silica gel 60 (0.040-0.063 mm) (Merck).
â-Glc-Py-γ-Py-Ind (2). Prepared as mentioned above for â-Gal-
Py-γ-Py-Ind 1, using a solution of (AcO)4- â-Glc-Py-γ-Py-Ind 2(OAc)
Calf thymus DNA (ct-DNA) was purchased from Sigma (Lot
90k7063) and used without further purification. O-Methyl-R-D-glu-
(52) Vicente, V.; Martin, J.; Jime´nez-Barbero, J.; Chiara, J.-L.; Vicent, C.
(53) Marky, L. A.; Breslauer, K. J. Proc. Natl. Acad. Sci. 1987, 84, 4359-
Chem.sEur. J. 2004, 10, 4240-4251.
4363.
9
9528 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
(100 mg, 0.122 mmol) in MeOH (10 mL) and 2 mL of a solution of
sodium (100 mg, 4.34 mmol) in MeOH (10 mL) to afford the title
compound as a yellow film (33.0 mg, 42%). An analytical sample was
prepared by precipitation from a solution in CHCl3 (10 mL) and MeOH
(5 mL) with Et2O, which gave 2 as an amorphous white solid. Mp )
118.1 (CH), 118.5 (CH), 119.7 (CH), 121.5 (CH), 121.6 (C), 121.9
(C), 122.2 (C), 123.2 (C), 123.3 (CH), 127.1 (C), 131.6 (C), 136.5
(C), 158.2 (C), 161.2 (C), 161.3 (C), 169.2 (C). MS (ES+) m/z (%):
1243 (5.8) [2M + H]+, 644 (19.2) [M + Na]+, 622 (100) [M + H]+
]+. Elemental analysis calcd (%) for C30H35N7O8 (621.64): C, 57.96;
H, 5.67; N, 15.77. Found: C, 57.89; H, 5.68; N, 15.69.
196-197 °C. IR (KBr): ν ) 3600-2800, 1638, 1584, 1532 cm-1
.
[R]D27 ) +12.9 (c ) 0.760 in [D6] DMSO). 1H NMR (200 MHz, [D6]
DMSO): δ ) 1.79 (m, 2H), 2.29 (m, 2H), 3.21 (m, 2H), 3.64 (m,
1H), 3.79 (s, 3H), 3.84 (s, 3H), 4.51 (t, J ) 5.7 Hz, 1H), 4.81 (d, J )
5.3 Hz, 1H), 4.87 (d, J ) 4.4 Hz, 1H), 4.96 (d, J ) 4.0 Hz, 1H), 6.85
(d, J ) 1.7 Hz, 1H), 6.90 (d, J ) 1.7 Hz, 1H), 7.06 (m, 1H), 7.19 (m,
2H), 7.28 (s, 2H), 7.46 (d, J ) 8.2 Hz, 1H), 7.65 (d, J ) 7.9 Hz, 1H),
8.13 (t, J ) 4.9 Hz, 1H), 8.36 (d, J ) 8.8 Hz, 1H), 9.82 (s, 1H), 10.29
(s, 1H), 11.60 (s, 1H). 13C NMR (200 MHz, DMSO): 25.6, 33.2, 33.5,
35.8, 36.0, 61.0, 70.1, 71.8, 77.6, 78.4, 79.4, 79.6, 103.2, 104.1, 112.5,
118.2, 119.6, 121.5, 121.5, 121.8, 122.2, 123.2, 123.2, 127.0, 127.0,
131.6, 161.1, 161.1, 169.6. MS (ES+) m/z (%): 1325 (9) [2M + Na]+,
1304 (3) [2M + Na]+, 674 (90) [M + Na]+, 652 (100) [M + H]+.
Elemental analysis calcd (%) for C31H37N7O9 (651.7): C, 57.14; H,
5.72; N, 15.05. Found: C, 57.19; H, 5.68; N, 15.1.
The R anomer of Xyl-Py-γ-Py-Ind: Rf ) 0.59 (EtOAc/MeOH 7:3
v/v). [R]D23 ) -7.8 (c ) 0.11 in methanol). 1H NMR (400 MHz, [D6]
DMSO): δ ) 1.79 (m, 2H, CH2b), 2.28 (t, J ) 7.4 Hz, 2H, CH2a),
3.21 (m, 2H, CH2c), 3.30-3.76 (m, 5H, 5CH), 3.77 (s, 3H, CH3A),
3.83 (s, 3H, CH3B), 4.98 (s, 1H, OH), 5.04 (s, 1H, OH), 5.17 (s, 1H,
OH), 5.36 (dd, J ) 8.4 Hz, 3.2 Hz, 1H, CH1), 6.80 (d, J ) 1.6 Hz,
1H, CH), 6.89 (d, J ) 1.6 Hz, 1H, CH), 7.05 (dd, J ) 7.2 Hz, 8.0 Hz,
1H, CH), 7.17-7.21 (m, 2H, 2CH), 7.28 (d, J ) 1.2 Hz, 2H, 2CH),
7.45 (d, J ) 8.4 Hz, 1H, CH), 7.64 (d, J ) 8.0 Hz, 1H, CH), 7.70 (d,
J ) 8.8 Hz, 1H, NH5), 8.11 (t, J ) 5.6 Hz, 1H, NH3), 9.81 (s, 1H,
NH4), 10.32 (s, 1H, NH2), 11.65 (s, 1H, NH1). 13C NMR (300 MHz,
[D6] DMSO): δ ) 21.7 (CH2), 33.3 (CH2), 36.1 (2CH3), 38.2 (CH2),
64.9 (CH2), 68.8 (CH), 70.5 (CH), 70.6 (CH), 75.8 (CH), 102.9 (CH),
104.1 (CH), 104.6 (CH), 112.3 (CH), 118.1 (CH), 118.6 (C), 119.8
(CH), 121.5 (CH), 121.6 (CH), 122.0 (C), 122.1 (C), 123.2 (C), 123.4
(CH), 127.1 (C), 131.7 (C), 136.6 (C), 158.2 (CO), 161.1 (CO), 161.2
(CO), 169.3 (CO). MS (ES+) m/z (%): 1265 (14) [2M + Na]+, 644
(41) [M + Na]+, 622 (100) [M + H]+. Elemental analysis calcd (%)
for C30H35N7O8 (621.64): C, 57.96; H, 5.67; N, 15.77. Found: C, 57.99;
H, 5.75; N, 15.81.
(AcO)4-â-Glc-Py-γ-Py-Ind [2(OAc)]. Prepared as mentioned above
for 1(OAc), using a solution of (AcO)4-â-Glc-Py-NO2 10 (500 mg,
1.00 mmol). Compound (AcO)4-Glc-Py-γ-Py-Ind 2(OAc) obtained as
an amorphous white solid (450 mg, 55%). Mp ) 160-161 °C; IR
(KBr): ν ) 1749, 1645, 1234 cm-1. [R]D25 ) +10.6 (c ) 1.56 in [D6]
1
DMSO). H NMR (400 MHz, [D6] DMSO): δ ) 1.80 (m, 2H), 1.90
(s, 3H), 1.94 (s, 3H), 1.99 (s, 3H), 2.00 (s, 3H), 2.28 (t, J ) 7.4 Hz,
2H), 3.23 (m, 2H), 3.79 (s, 3H), 3.84 (s, 3H), 4.00 (dd, J ) 1.9 Hz,
12.3 Hz, 1H), 4.10 (ddd, J ) 2.1 Hz, 4.3 Hz, 10.0 Hz, 1H), 4.19 (dd,
J ) 4.5 Hz, 12.4 Hz, 1H), 4.91 (dd, J ) 9.7 Hz, 9.9 Hz, 1H), 5.09 (dd,
J ) 9.3 Hz, 9.5 Hz, 1H), 5.36 (dd, J ) 9.5 Hz, 9.7 Hz, 1H), 5.50 (dd,
J ) 9.3 Hz, 9.5 Hz, 1H), 6.77 (d, J ) 2.0 Hz, 1H), 6.90 (d, J ) 1.8
Hz, 1H), 7.06 (m, 1H), 7.17-7.23 (m, 2H), 7.28 (m, 2H), 7.46 (dd, J
) 0.9 Hz, 8.2 Hz, 1H), 7.65 (d, J ) 7.9 Hz, 1H), 8.09 (dd, J ) 5.5 Hz,
5.9 Hz, 1H), 8.66 (d, J ) 9.5 Hz, 1H), 9.84 (s, 1H), 10.27 (s, 1H),
11.59 (d, J ) 1.8 Hz, 1H). 13C NMR (100 MHz, [D6] DMSO): δ )
20.3, 20.3, 20.5, 25.6, 33.2, 36.0, 36.1, 38.2, 61.7, 67.9, 70.6, 72.0,
73.1, 77.2, 79.1, 102.8, 104.1, 104.9, 112.3, 118.1, 119.2, 119.8, 121.4,
121.5, 121.6, 122.2, 123.3, 123.4, 127.1, 127.1, 136.6, 158.2, 160.9,
161.2, 169.0, 169.3, 169.3, 169.5, 170.0. MS (ES+) m/z (%): 1662
(1) [2M + H]+, 842 (22) [M + Na]+, 820 (79) [M + H]+, 410 (100)
[(AcO)3-Glc-Py-NH-]. Elemental analysis calcd (%) for C39H45N7O13
+ 1.2H2O (841.4): C, 55.67; H, 5.68; N, 11.65. Found: C, 55.42; H,
5.49; N, 11.44.
(AcO)3-â-Xyl-Py-γ-Py-Ind [3(OAc)]. (AcO)3-â-Xyl-Py-γ-Py-Ind
3(OAc) was prepared as described above for 1(OAc) using a solution
of a mixture of R and â anomers of (AcO)3-Xyl-Py-NO2 (365 mg,
0.854 mmol) in CH2Cl2 (10 mL) with 10% Pd-C (45 mg) (H2, atm
pressure, 20 h). A mixture of both anomers of (AcO)4-Xyl-Py-γ-Py-
Ind as a gum was contained (R:â ) 1:2) (269 mg, 48%). An analytical
sample of the â-isomer was separated and characterized.
(AcO)3-â-Xyl-Py-γ-Py-Ind 3(OAc): Rf ) 0.58 (EtOAc/MeOH 9:1
1
v/v); H NMR (200 MHz, [D6] DMSO): δ ) 1.79 (m, 2H, CH2b),
1.91 (s, 3H, CH3), 1.97 (s, 3H, CH3), 2.00 (s, 3H, CH3), 2.27 (m, 2H,
CH2a), 3.21 (m, 2H, CH2c), 3.56 (m, 1H, CH), 3.77 (s, 3H, CH3A),
3.83 (s, 3H, CH3B), 3.89 (m, 1H, CH), 4.81 (m, 1H, CH), 5.06 (dd, J
) 9.2 Hz, 9.4 Hz, 1H, CH1), 5.30 (m, 2H, 2CH), 6.72 (d, J ) 1.4 Hz,
1H, CH), 6.89 (d, J ) 1.8 Hz, 1H, CH), 7.04 (dd, J ) 7.8 Hz, 7.2 Hz,
1H, CH), 7.20 (m, 2H, 2CH), 7.27 (s, 2H, 2CH), 7.45 (d, J ) 8.0 Hz,
1H, CH), 7.64 (d, J ) 7.8 Hz, 1H, CH), 8.10 (m, 1H, NH3), 8.58 (d,
J ) 9.2 Hz, 1H, NH5), 9.85 (s, 1H, NH4), 10.28 (s, 1H, NH2), 11.60
(s, 1H, NH1). 13C NMR (200 MHz, [D6] DMSO): δ ) 20.2 (3CH3),
25.5 (CH2), 33.2 (CH2), 35.8 (2CH3), 38.1 (CH2), 63.2 (CH), 68.6 (CH),
70.6 (CH), 72.7 (CH), 77.6 (CH), 102.7 (CH), 104.1 (CH), 104.8 (CH),
112.1 (CH), 118.0 (CH), 118.9 (CH), 119.6 (CH), 121.5 (CH), 121.9
(3C), 122.0 (C), 123.2 (CH), 127.0 (C), 131.6 (C), 136.5 (C), 158.1
(HNCO), 160.9 (HNCO), 161.1 (HNCO), 168.9 (COCH3), 169.2
(COCH3), 169.4 (COCH3). MS (ES+) m/z (%): 1517 (8) [2M + Na]+,
1495 (10) [2M + H]+, 770 (30) [M + Na]+, 748 (100) [M + H]+.
â-Xyl-Py-γ-Py-Ind (3). Prepared as described above for 1 using a
solution of (AcO)3-â-Xyl-Py-γ-Py-Ind 3 (OAc) (100 mg, 0.133 mmol)
in MeOH (10 mL) and 2 mL of a solution of sodium (100 mg, 4.34
mmol) in MeOH (10 mL). The product was purified by column
chromatography (EtOAc/MeOH 8:2) and afforded three fractions. The
R anomer of Xyl-Py-γ-Py-Ind (15.5 mg, 19%), the â anomer of Xyl-
Py-γ-Py-Ind 3 (27.2 mg, 33%), and a mixture of both anomer R/â of
Xyl-Py-γ-Py-Ind (23.3 mg, 29%) were obtained.
â-Xyl-Py-γ-Py-Ind 3: Rf ) 0.40 (EtOAc/MeOH 7:3 v/v). [R]D
)
The R anomer of (AcO)3-Xyl-Py-γ-Py-Ind could not be purified.
The mixture of R and â anomers (AcO)3-Xyl-Py-γ-Py-Ind showed
the 1H NMR as follows: 1H NMR (200 MHz, [D6] DMSO): δ ) 1.79
(m, 4H, CH2bR, CH2bâ), 1.91 (s, 3H, CH3â), 1.97 (s, 3H, CH3â), 1.98
(s, 3H, CH3R), 2.00 (s, 3H, CH3â), 2.01 (s, 3H, CH3R), 2.04 (s, 3H,
CH3R), 2.28 (m, 4H, CH2cR, CH2câ), 3.23 (m, 4H, CH2aR, CH2aâ),
22
+5.5 (c ) 1 in DMSO). 1H NMR (400 MHz, [D6] DMSO): δ ) 1.79
(m, 2H, CH2), 2.28 (t, J ) 7.6 Hz, 2H, CH2), 3.04 (t, J ) 10.8 Hz, 2H,
CH2), 3.10-3.36 (m, 4H, 4CH), 3.65 (m, 1H, CH), 3.77 (s, 3H, CH3),
3.83 (s, 3H, CH3), 4.77 (t, J ) 8.8 Hz, 1H, CH1), 4.86 (d, J ) 5.6 Hz,
1H, OH), 4.95 (d, J ) 5.2 Hz, 1H, OH), 5.04 (d, J ) 4.8 Hz, 1H, OH),
6.83 (d, J ) 1.6 Hz, 1H, CH), 6.89 (d, J ) 2.0 Hz, 1H, CH), 7.05 (dd,
J ) 6.8 Hz, 8.0 Hz, 1H, CH), 7.17-7.21 (m, 2H, 2CH), 7.28 (d, J )
1.6 Hz, 2H, 2CH), 7.45 (d, J ) 8.4 Hz, 1H, CH), 7.64 (d, J ) 8.0 Hz,
1H, CH), 8.12 (t, J ) 5.6 Hz, 1H, NH3), 8.35 (d, J ) 8.8 Hz, 1H,
NH5), 9.83 (s, 1H, NH4), 10.31 (s, 1H, NH2), 11.62 (s, 1H, NH1).
13C NMR (500 MHz, [D6] DMSO): δ ) 26.6 (CH2), 33.2 (CH2), 35.9
(CH3py), 36.1 (CH3py), 38.2 (CH2), 67.3 (CH2), 69.7 (CH), 71.6 (CH),
77.6 (CH), 80.5 (CH), 102.8 (CH), 104.1 (CH), 104.6 (CH), 112.3 (CH),
P
P
3.56 (m, 2H, 2CHâ), 3.75 (s, 3H, CH3 R), 3.77 (s, 3H, CH3 â), 3.83
P
P
(s, 6H, CH3 R, CH3 â), 4.83 (m, 5H, 4CHR, CHâ), 5.05 (t, J ) 9.4
Hz, 1H, CHâ), 5.32 (m, 2H, 2CHâ), 5.74 (m, 2H, 2CHR), 6.72 (d, J
) 1.8 Hz, 2H, CHPR, CHPâ), 6.89 (d, J ) 1.6 Hz, 2H, CHPR, CHPâ),
6.95 (d, J ) 1.8 Hz, 2H, CHR, CHâ), 7.05 (dd, J ) 7.0 Hz, 6.8 Hz,
2H, CHR, CHâ), 7.20 (m, 4H, 4CHRâ), 7.28 (s, 2H, CHR, CHâ), 7.45
(d, J ) 8.2 Hz, 2H, CHR, CHâ), 7.65 (d, J ) 8.2 Hz, 2H, CHR, CHâ),
8.11 (br, 2H, NH3R, NH3â), 8.59 (d, J ) 9.6 Hz, 1H, NH5â), 8.79 (d,
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9529

A R T I C L E S
Martin et al.
J ) 9.2 Hz, 1H, NH5R), 9.85 (s, 2H, NH4R, NH4â), 10.29 (s, 2H,
NH2R, NH2â), 11.61 (s, 2H, NH1R, NH1â).
Elemental analysis calcd (%) for C19H20N4O4 (368.4): C, 61.95; H,
5.47; N, 15.21. Found: C, 61.67; H, 5.31; N, 15.40.
MeNH-γ-Py-Ind (8). The activation reaction of HO-γ-Py-Ind 7
(300 mg, 0.844 mmol) with DIPC (119 µL, 0.757 mmol) and HOBt
(114 mg, 0.757 mmol) in DMF (5 mL) (room temperature, 24 h) was
followed by addition of methylamine hydrochloride (102 mg, 1.51
mmol) and Et3N (422 µL, 3.03 mmol) in DMF (5 mL) (24 h).
Purification of the residue by column chromatography (SiO2, 1-10%
MeOH in CH2Cl2) afforded the title compound as an off-white
amorphous solid (230 mg, 80%). Trituration of this solid with MeOH
and then Et2O afforded analytically pure 8 (140 mg). Mp ) 240-242
°C. IR (KBr): ν ) 1645, 1631, 1579, 1570 cm-1. 1H NMR (200 MHz,
[D6] DMSO): δ ) 1.71 (m, 2H), 2.10 (m, 2H), 2.56 (d, J ) 4.4 Hz,
3H), 3.17 (m, 2H), 3.83 (s, 3H), 6.87 (d, J ) 2.0 Hz, 1H), 7.05 (dd, J
) 7.4 Hz, 7.6 Hz, 1H), 7.19 (dd, J ) 7.4 Hz, 7.0 Hz, 1H), 7.29 (d, J
) 1.4 Hz, 2H), 7.45 (d, J ) 8.0 Hz, 1H), 7.66 (m, 1H), 7.79 (m, 1H),
8.11 (dd, J ) 5.8 Hz, 5.6 Hz, 1H), 10.33 (br s, 1H), 11.64 (br s, 1H).
13C NMR (50 MHz, [D6] DMSO): δ ) 25.3, 25.4, 32.9, 35.8, 38.2,
102.8, 104.0, 112.2, 118.0, 119.7, 121.4, 121.5, 123.3, 123.3, 127.1,
131.6, 136.5, 158.2, 161.1, 172.2. MS (ES+) m/z (%): 785 (15) [2M
+ Na]+, 763 (9) [2M + H]+, 404 (5) [M + Na]+, 382 (100) [M +
H]+. Elemental analysis calcd (%) for C20H23N5O3 (381.43): C, 62.98;
H, 6.08; N, 18.36. Found: C, 62.89; H, 6.38; N, 18.65.
â-Gal-Py-NHAc (4). A solution of (AcO)4-â-Gal-Py-NHAc 17
(80.0 mg, 0.156 mmol) in MeOH (5 mL) was added to a solution of
sodium (100 mg, 4.34 mmol) in MeOH (10 mL) to afford the title
compound as a colorless film that crystallizes from the crude mixture
by Et2O addition (44.0 mg, 82%). Mp ) 148-149 °C; [R]D22 ) -2.6
(c ) 1 in H2O). 1H NMR (400 MHz, [D6] DMSO): δ ) 1.95 (s, 3H),
3.30-3.44 (m, 3H), 3.49 (m, 1H), 3.59 (dd, J ) 9.2 Hz, 9.2 Hz, 1H),
3.70 (d, J ) 3.1 Hz, 1H), 3.78 (s, 3H), 4.83 (m, 1H), 6.84 (d, J ) 1.8
Hz, 1H), 7.16 (d, J ) 1.8 Hz, 1H), 8.36 (d, J ) 9.2 Hz, 1H), 9.80 (s,
1H); 13C NMR (300 MHz, [D6] DMSO): δ ) 23.0, 36.3, 54.9, 60.5,
68.3, 69.1, 74.4, 76.7, 80.1, 104.4, 118.4, 122.0, 161.3, 166.5. MS
(ES+) m/z (%): 709 (34) [2M + Na]+, 687 (19) [2M + H]+, 453
(37), 366 (19) [M + Na]+, 344 (100) [M + H]+, 182 (18).
â-Glc-Py-NHAc (5). Prepared as described above for 4, from a
solution of (AcO)4-â-Glc-Py-NHAc 18 (110 mg, 0.215 mmol) in
MeOH. Precipitation from the MeOH solution with Et2O afforded an
analytical sample of the title compound as an amorphous white solid
(61 mg, 83%). Mp ) 161-163 °C; IR (KBr): ν ) 3600-3200, 1644
cm-1. [R]D27 ) + 18.6 (c ) 1.37 in [D6] DMSO). 1H NMR (300 MHz,
[D6] DMSO): δ ) 1.96 (s, 3H), 3.04-3.15 (m, 2H), 3.21 (dd, J ) 8.1
Hz, 8.8 Hz, 1H), 3.31 (dd, J ) 8.8 Hz, 9.0 Hz, 1H), 3.40-3.67 (m,
2H), 3.79 (s, 3H), 4.88 (dd, J ) 8.8 Hz, 9.0 Hz, 1H), 6.83 (d, J ) 1.8
Hz, 1H), 7.14 (d, J ) 1.8 Hz, 1H), 8.33 (d, J ) 8.9 Hz, 1H), 9.77 (s,
1H). 13C NMR (50 MHz, [D6] DMSO): δ ) 22.9, 36.0, 61.0, 70.1,
71.8, 77.6, 78.4, 79.6, 104.5, 118.3, 121.9, 122.2, 161.1, 166.4. MS
(ES+) m/z (%): 739 (21) [2M + K]+, 709 (100) [2M + H]+, 374 (18)
[M + K]+, 366 (28) [M + Na]+, 344 (73) [M + H]+, 182 (25).
â-Xyl-Py-NHAc (6). Prepared as described above for 4, from a
solution of (AcO)3-â-Xyl-Py-NHAc 19 (96.0 mg, 0.218 mmol) in
MeOH. The solution was acidified to pH ) 6 with IR-120 ion-exchange
resin, and the resin was removed by filtration. Evaporation of the solvent
under reduced pressure and purification by column chromatography
(SiO2, EtOAc/MeOH 7:3) afforded â-Xyl-Py-NHAc 6 (50.2 mg, 74%)
as a white powder. Rf ) 0.61 (EtOAc/MeOH 1:1 v/v). [R]D22 ) +10.77
(c ) 0.8 in DMSO). 1H NMR (200 MHz, [D6] DMSO): δ ) 1.95 (s,
3H, CH3), 3.00-4.00 (m, 5H, 5CH), 3.78 (s, 3H, CH3), 4.77 (dd, J )
9.2 Hz, 8.6 Hz, 1H, CH1), 4.89 (d, J ) 4.8 Hz, 1H, OH), 4.96 (d, J )
3.8 Hz, 1H, OH), 5.06 (d, J ) 2.6 Hz, 1H, OH), 6.81 (s, 1H, CH),
7.14 (s, 1H, CH), 8.35 (d, J ) 8.6 Hz, 1H, NH5), 9.80 (s, 1H, NH4).
13C NMR (400 MHz, [D6] DMSO): δ ) 23.0 (CH3), 36.2 (CH3), 67.4
(CH2), 69.7 (CH), 71.6 (CH), 77.7 (CH), 80.5 (CH), 104.4 (CH), 118.4
(CH), 122.0 (C), 122.2 (C), 161.3 (HNCO), 166.5 (HNCO). MS (ES+)
m/z (%): 649 (59) [2M + Na]+, 336 (34) [M + Na]+, 314 (100) [M
+ H]+. Elemental analysis calcd (%) for C13H19O6N3 (313.13): C,
49.84; H, 6.11; N, 13.41. Found: C, 49.57; H, 6.40; N, 13.21.
(AcO)4-â-Gal-Py-NO2 (9). A mixture of 1-methyl-4-nitropyrrole-
2-carboxylic acid 1341 (108 mg, 0.634 mmol) was treated with thionyl
chloride (560 µL, 7.68 mmol) and heated to reflux for 1 h. The mixture
was cooled to room temperature, and an excess of reagent was removed
under moderate vacuum (∼80 mmHg) to afford the crude acid chloride
1442,43 as an off-white solid. Separately, the amine 2,3,4,6-tetra-O-acetyl-
â-D-galactopyranosylamine 1254 was prepared (200 mg, 0.575 mmol)
from the corresponding azide. The amine freshly prepared was
immediately added to the crude acid chloride in anhydrous CH2Cl2 (5
mL) in the presence of Et3N (80.0 µL, 0.575 mmol). The mixture was
heated to reflux for 5 h, before cooling to room temperature overnight.
Then the solvent was removed at reduced pressure. The residue was
purified by column chromatography (SiO2, 0.5% MeOH in CH2Cl2) to
afford the title compound 9 as an off-white amorphous solid (250 mg,
87%). An analytical sample was prepared by recrystallization from
ethanol. Mp ) 180-182 °C. IR (KBr): ν ) 1750, 1729, 1688, 1537,
1
1368, 1315 cm-1. [R]D ) +3.29 (c ) 2.47 in acetone). H NMR
23
(200 MHz, [D6] DMSO): δ ) 1.92 (s, 6H), 1.99 (s, 3H), 2.13 (s, 3H),
3.90 (s, 3H), 4.02 (m, 2H), 4.36 (t, J ) 6.1 Hz, 1H), 5.36-5.21 (m,
3H), 5.50 (t, J ) 9.2 Hz, 1H), 7.58 (d, J ) 2.0 Hz, 1H), 8.18 (d, J )
2.0 Hz, 1H), 9.21 (br d, J ) 9.5 Hz, 1H). 13C NMR (100 MHz, [D6]
DMSO): δ ) 20.4, 20.5, 20.6 (fourth acetate signal not observed),
37.7, 61.4, 67.6, 68.2, 71.0, 71.4, 77.3, 109.1, 125.0, 128.8, 133.9, 159.9,
169.2, 169.5, 170.0, 170.0. MS (ES+) m/z (%): 1021 (4) [2M + Na]+,
522 (100) [M + Na]+, 500 (82) [M + H]+, 331 (78) [M-(NHCO-Py-
NO2)]+, 128 (59). Elemental analysis calcd (%) for C20H25N3O12
(499.4): C, 48.10; H, 5.05; N, 8.41. Found: C, 48.40; H, 5.35; N,
8.56.
HO-γ-Py-Ind (7). A solution of EtO-γ-Py-Ind 23 from the previous
coupling reaction [13.6 g of a 2.54 equiv of 23 (9.76 g, 24.6 mmol)
and 1 equiv of diisopropylurea] in MeOH (850 mL) were treated with
aqueous NaOH (1 M, 525 mL). An instantaneous yellow color
corresponded with reaction to completion, and after filtration at reduced
pressure the filtrate was acidified to pH ) 1 with aqueous HCl (1 M).
A sticky, biscuit-colored aggregate was collected by filtration at reduced
pressure and dried under vacuum at 40 °C overnight to afford the title
compound as an off-white amorphous solid (7.52 g, 83%). Mp ) 139-
142 °C. IR (KBr): ν ) 3600-2700, 1713, 1640 cm-1. 1H NMR (200
MHz, CDCl3): δ ) 1.73 (m, 2H), 2.27 (t, J ) 7.1 Hz, 2H), 3.24 (m,
2H), 3.84 (s, 3H), 6.90 (d, J ) 1.8 Hz, 1H), 7.05 (m, 1H), 7.23 (m,
2H), 7.31 (s, J ) 1.6 Hz, 1H), 7.47 (d, J ) 8.2 Hz, 1H), 7.65 (d, J )
7.9 Hz, 1H), 8.14 (t, J ) 5.5 Hz, 1H), 10.38 (br s, 1H), 11.67 (br s,
1H). 13C NMR (125 MHz, CDCl3): δ ) 24.9, 31.3, 36.1, 38.0, 103.0,
104.2, 112.4, 118.2, 119.9, 121.7, 121.8, 123.3, 123.5, 127.2, 131.8,
136.7, 158.3, 161.4, 174.5. MS (ES+) m/z (%): 759 (77) [2M + Na]+,
391 (34) [M + Na]+, 369 (77) [M + H]+, 167 (55), 145 (100).
(AcO)4-â-Glc-Py-NO2 (10). Prepared as for 9 described above using
carboxylic acid 13 (1.00 g, 5.88 mmol) and thionyl chloride (5.15 mL,
70.6 mmol) (reflux, 3 h). A solution of a freshly prepared 2,3,4,6-
tetra-O-acetyl-â-D-glucopyranosylamine 1555 (200 mg, 0.575 mmol)
with Et3N (80.0 µL, 0.575 mmol) in CH2Cl2 (20 mL) was added. The
reaction was stirred at room temperature overnight. Purification by
column chromatography (SiO2, EtOAc) afforded the title compound
10 as an off-white amorphous solid (2.37 g, 89%). An analytical sample
was prepared by recrystallization from EtOH/MeOH (20:1 v/v). Mp )
185-186 °C. IR (KBr): ν ) 1750, 1530, 1314, 1232, cm-1. [R]D
)
24
-36.7 (c ) 1.64 in acetone). 1H NMR (200 MHz, [D6] DMSO): δ )
(54) Christiansen-Brams, I.; Meldal, M.; Bock, K. J. Chem. Soc., Perkin Trans.
1 1993, 1461-1471.
(55) Takeda, T.; Sugiura, Y.; Ogihara, Y.; Shibata, S. Can. J. Chem. 1980, 58,
2600-2603.
9
9530 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
1.92 (s, 3H), 1.95 (s, 3H), 2.00 (s, 6H), 3.91 (s, 3H), 3.96-4.23 (m,
3H), 4.92 (dd, J ) 9.7, 9.9 Hz, 1H), 5.03 (dd, J ) 9.3, 9.5 Hz, 1H),
5.40 (dd, J ) 9.5, 9.5 Hz, 1H), 5.56 (dd, J ) 9.3, 9.3 Hz, 1H), 7.53 (d,
J ) 1.9 Hz, 1H), 8.18 (d, J ) 1.9 Hz, 1H), 9.14 (d, J ) 9.3 Hz, 1H).
13C NMR (50 MHz, [D6] DMSO): δ ) 20.2, 20.3, 20.5 (fourth acetate
signal not observed), 37.3, 61.6, 67.8, 70.5, 72.1, 72.9, 77.0, 106.7,
124.9, 128.6, 133.8, 159.8, 168.9, 169.2, 169.4, 169.8. MS (ES+) m/z
(%): 1021 (2) [2M + Na]+, 522 (27) [M + Na]+, 500 (100) [M +
H]+, 412 (27), 390 (32), 348 (55), 331 (23). Elemental analysis calcd
(%) for C20H25N3O12 (499.4): C, 48.10; H, 5.05; N, 8.41. Found: C,
47.96; H, 5.22; N, 8.32.
1.34 in [D6] DMSO). 1H NMR (300 MHz, [D6] DMSO): δ ) 1.88 (s,
3H), 1.93 (s, 3H), 1.94 (s, 3H), 1.98 (s, 6H), 3.77 (s, 3H), 3.98 (dd, J
) 1.8 Hz, 12.2 Hz, 1H), 4.07-4.18 (m, 2H), 4.92 (t, J ) 9.3 Hz, 1H),
5.07 (t, J ) 9.0 Hz, 1H), 5.35 (t, J ) 9.6 Hz, 1H), 5.50 (t, J ) 9.3 Hz,
1H), 6.72 (d, J ) 1.8 Hz, 1H), 7.17 (d, J ) 1.8 Hz, 1H), 8.64 (d, J )
9.5 Hz, 1H), 9.79 (s, 1H). 13C NMR (50 MHz, [D6] DMSO): δ )
20.2, 20.2, 20.2, 20.4, 22.8, 35.9, 61.7, 67.9, 70.6, 72.0, 73.1, 77.1,
104.8, 118.9, 121.4, 122.1, 160.8, 166.4, 168.8, 169.1, 169.3, 169.8.
MS (ES+) m/z (%): 1045 (31) [2M + Na]+, 1023 (38) [2M + H]+,
542 (25) [M + K]+, 534 (34) [M + Na]+, 512 (100) [M + H]+.
Elemental analysis calcd (%) for C22H29N3O11 (511.48): C, 51.66; H,
5.71; N, 8.22. Found: C, 51.38; H, 5.63; N, 8.04.
(AcO)3-â-Xyl-Py-NO2 (11). 2,3,4-Tri-O-acetyl-â-D-xylopyranosyl-
amine 16 obtained by hydrogenation of the correspondent azide (1.01
g, 3.35 mmol) was added to a mixture of HO-Py-NO2 13 (525 mg,
3.09 mmol), HOBt (626 mg, 4.63 mmol), and DIPC (725 µL, 4.63
mmol) dissolved in DMF (4.0 mL) previously stirred at room
temperature for 24 h. After 16 h, the residue was washed with aqueous
sodium hydrogen carbonate then with water, dried over Na2SO4, and
filtered. After solvent evaporation, the residue was purified by column
chromatography (SiO2, CH2Cl2/EtOAc 8:2) to afford (AcO)3-Xyl-Py-
NO2 (1.0 g, 70%) as a gum containing a mixture of R and â isomers
(R:â ) 1:5). An analytical sample of the â-isomer was separated and
(AcO)3-â-Xyl-Py-NHAc (19). A mixture of R and â anomers of
(AcO)3-Xyl-Py-NO2 (550.0 mg, 1.28 mmol) dissolved in MeOH (10
mL) and CH2Cl2 (5 mL) was treated with 10% Pd/C (68.5 mg, 5 mol
%) and hydrogenated overnight. The mixture was filtered over a pad
of Celite, and the solvent was removed under reduced pressure. The
residue was taken up in CH2Cl2 (10 mL), and acetic anhydride (280
µL, 3.14 mmol) and Et3N (435 µL, 3.14 mmol) were added and stirred
at room temperature overnight. The mixture was evaporated under
reduced pressure and purified by column chromatography (EtOAc
100%) to afford a mixture of R and â anomers (R:â ) 1:5) of (AcO)3-
Xyl-Py-NHAc (394 mg, 72%). An analytical sample of each isomer
was separated and characterized. (AcO)3-â-Xyl-Py-NHAc 19: Rf ) 0.26
(EtOAc 100%). [R]D30 ) -52.5 (c ) 1 in chloroform). 1H NMR (200
MHz, CDCl3): δ ) 2.07, 2.05 (s, 9H, 3CH3), 2.11 (s, 3H, CH3), 3.48
(dd, J ) 11.4 Hz, 10.6 Hz, 1H, CH), 3.88 (s, 3H, CH3), 4.09 (dd, J )
6.0 Hz, 11.5 Hz, 1H, CH), 4.95 (dd, J ) 9.4 Hz, 9.6 Hz, 1H, CH),
5.02 (m, 1H, CH), 5.18 (dd, J ) 9.0 Hz, 9.2 Hz, 1H, CH), 5.35 (dd, J
) 9.4 Hz, 9.6 Hz, 1H, CH1), 6.32 (d, J ) 1.8 Hz, 1H, CH), 6.73 (d,
J ) 9.0 Hz, 1H, NH5), 7.16 (s, 1H, NH4), 7.29 (d, J ) 1.8 Hz, 1H,
CH). 13C NMR (200 MHz, CDCl3): δ ) 20.5 (CH3), 20.6 (CH3), 20.7
(CH3), 23.4 (CH3), 36.7 (CH3), 64.3 (CH), 69.1 (CH), 70.9 (CH), 72.2
(CH), 79.0 (CH1), 104.3 (C), 120.9 (C), 121.5 (C), 121.6 (C), 161.3
(HNCO), 167.6 (HNCO), 169.9 (COCH3), 170.0 (COCH3), 171.6
(COCH3). MS (ES+) m/z (%): 901 (49) [2M + Na]+, 879 (33.5) [2M
+ H]+, 462 (15.4) [M + Na]+, 440 (100) [M + H]+. Elemental analysis
calcd (%) for C19H25O9N3 (439.42): C, 51.93; H, 5.73; N, 9.56.
Found: C, 51.68; H, 5.69; N, 9.37. (AcO)3-r-Xyl-Py-NHAc: Rf )
0.20 (EtOAc 100%). [R]D28 ) - 4.09 (c ) 1 in chloroform). 1H NMR
(200 MHz, CDCl3): δ ) 2.11 (s, 3H, CH3), 2.13 (s, 3H, CH3), 2.14 (s,
3H, CH3), 2.17 (s, 3H, CH3), 3.88 (s, 3H, CH3), 4.00 (m, 2H, 2CH),
4.76 (m, 1H, CH), 4.87 (dd, J ) 2.7 Hz, 4.3 Hz, 1H, CH), 5.24 (t, J
) 4.2 Hz, 1H, CH), 5.70 (dd, J ) 8.7 Hz, 2.7 Hz, 1H, CH1), 6.52 (d,
J ) 2.2 Hz, 1H, CH), 6.65 (d, J ) 8.8 Hz, 1H, NH5), 7.14 (d, J ) 1.8
Hz, 1H, CH), 7.20 (s, 1H, NH4). 13C NMR (200 MHz, CDCl3): δ )
20.7 (CH3), 20.8 (2CH3), 23.4 (CH3), 36.7 (CH3), 63.8 (CH2), 66.7
(CH), 67.5 (CH), 68.6 (CH), 74.5 (CH), 104.6 (C), 120.3 (C), 121.5
(C), 121.8 (C), 161.1 (HNCO), 167.7 (HNCO), 169.1 (COCH3), 169.6
(COCH3), 169.8 (COCH3). MS (ES+) m/z (%): 879 (3) [2M + H]+,
462 (100) [M + Na]+, 440 (65) [M + H]+. Elemental analysis calcd
(%) for C19H25O9N3 (439.42): C, 51.93; H, 5.73; N, 9.56. Found: C,
51.58; H, 6.02; N, 9.18.
characterized. (AcO)3-â-Xyl-Py-NO2 11: Rf ) 0.39 (CH2Cl2/EtOAc
1
8:2 v/v). [R]D ) -53.9 (c ) 1 in chloroform). H NMR (200 MHz,
30
CDCl3): δ ) 2.05 (s, 3H, CH3), 2.10 (s, 3H, CH3), 2.11 (s, 3H, CH3),
3.47 (dd, J ) 10.8 Hz, 11.2 Hz, 1H, CH), 3.99 (s, 3H, CH3), 4.11 (dd,
J ) 5.4 Hz, 11.4 Hz, 1H, CH), 5.05 (m, 1H, CH), 5.07 (t, J ) 9.4 Hz,
1H, CH), 5.22 (dd, J ) 8.8 Hz, 9.2 Hz, 1H, CH1), 5.38 (dd, J ) 9.6
Hz, 9.4 Hz, 1H, CH), 7.19 (d, J ) 1.8 Hz, 1H, CH), 7.20 (d, J ) 8.6
Hz, 1H, NH), 7.58 (d, J ) 1.8 Hz, 1H, CH). 13C NMR (300 MHz,
CDCl3): δ ) 20.4 (CH3), 20.5 (CH3), 20.7 (CH3), 38.1 (CH3), 64.3
(CH5), 68.9 (CH), 71.2 (CH), 72.3 (CH), 78.5 (CH1), 108.7 (C), 124.8
(C), 127.5 (CH), 135.1 (CH), 160.4 (CONH), 169.9 (COCH3), 170.1
(COCH3), 171.8 (COCH3). MS (ES+) m/z (%): 687 (12.5) [2M +
Na]+, 450 (62.5) [M + Na]+, 428 (100) [M + H]+. Elemental analysis
calcd (%) for C17H21O10N3 (427.36): C, 47.78; H, 4.95; O, 37.44.
Found: C, 47.53; H, 4.97; N, 9.66.
(AcO)4-â-Gal-Py-NHAc (17). A solution of (AcO)4-â-Gal-Py-NO2
9 (400 mg, 0.801 mmol) in MeOH (20 mL) and EtOH (20 mL) was
treated with 10% Pd-C (85 mg, 10 mol %) and shaken under an
atmosphere of hydrogen at 45 psi overnight. The mixture was filtered,
and the solvent removed under reduced pressure. The residue was taken
up in CH2Cl2 (5 mL) and treated with a solution of acetyl chloride
(63.0 µL, 0.881 mmol) and Et3N (123 µL, 0.881 mmol) in CH2Cl2 (5
mL) dropwise and then stirred at room-temperature overnight. Evapora-
tion of the solvent and purification of the residue by repeated column
chromatography (SiO2, EtOAc) afforded the title compound as a glassy
film (83.0 mg, 20%). Mp ) 123-125 °C. IR (KBr): ν ) 1750, 1667,
1
1252, 1231 cm-1. [R]D ) - 7.05 (c ) 0.87 in acetone). H NMR
23
(200 MHz, CDCl3): δ ) 2.01 (s, 3H), 2.02 (s, 3H), 2.03 (s, 3H), 2.11
(s, 3H), 2.16 (s, 3H), 3.86 (s, 3H), 4.08-4.12 (m, 3H), 5.18-5.31 (m,
3H), 5.48 (br s, 1H), 6.40 (d, J ) 1.6 Hz, 1H), 6.87 (d, J ) 8.6 Hz,
1H), 7.29 (d, J ) 1.6 Hz, 1H), 7.68 (br s, 1H). 13C NMR (50 MHz,
CDCl3): δ ) 20.5, 20.6, 20.7, 20.9, 23.6, 36.8, 61.0, 67.2, 68.4, 70.8,
72.1, 78.8, 103.9, 120.8, 121.4, 121.5, 160.8, 167.4, 169.8, 170.1, 170.4,
171.9. MS (ES+) m/z (%): 1047 (10) [2M + Na]+, 1024 (14) [2M +
H]+, 513 (100) [M + H]+. Elemental analysis calcd (%) for C22H29N3O11
+ H2O (528.5): C, 49.90; H, 5.90; N, 7.93. Found: C, 50.20; H, 5.60;
N, 7.64.
EtO-γ-Py-NO2 (20). The title compound was prepared according
to the method of Xiao et al.44 with 1-methyl-4-nitro-2-trichloro-
acetylpyrrole (1.00 g, 3.91 mmol), Et3N (1.20 mL, 8.62 mmol), ethyl-
4-aminobutyrate hydrochloride (721 mg, 4.31 mmol), and EtOAc (20
mL) to afford the target ester after 24 h at room temperature. Then, it
was purified by column chromatography (SiO2, EtOAc) as yellow oil
(1.00 g, 90%) which slowly crystallized on standing. An analytical
sample was prepared by precipitation from ethanol at room temperature
overnight to afford an off-white amorphous solid. Mp ) 68-70 °C; Rf
) 0.53 (AcOEt/hexane 6:4 v/v). IR (KBr):44 ν ) 1728, 1666, 1526,
(AcO)4-â-Glc-Py-NHAc (18). Prepared as for 17 described above
from (AcO)4-â-Glc-Py-NO2 10 (500 mg, 1.00 mmol) and 10% Pd-C
(106 mg, 10 mol %) in MeOH (30 mL) (H2, rt, 18 h). The crude material
was purified by column chromatography (SiO2, 50% EtOAc/hexane)
and then precipitation from CHCl3 with hexane to afford the title
compound as a white amorphous solid (190 mg, 37%). Mp ) 120-
123 °C. IR (KBr): ν ) 1751, 1655, 1233 cm-1. [R]D27 ) -16.6 (c )
1
1318 cm-1. H NMR (200 MHz, CDCl3): δ ) 1.26 (t, J ) 7.2 Hz,
3H), 1.94 (t, J ) 6.6 Hz, 2H), 2.44 (t, J ) 6.6 Hz, 2H), 3.43 (q, J )
6.6 Hz, 2H), 3.98 (s, 3H), 4.16 (q, J ) 7.2 Hz, 2H), 6.59 (br s, 1H),
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9531

A R T I C L E S
Martin et al.
7.09 (d, J ) 1.8 Hz, 1H), 7.55 (d, J ) 1.8 Hz, 1H). 13C NMR (200
MHz, CDCl3): δ ) 14.0 (CH3), 24.2 (CH2), 31.8 (CH2), 37.8 (CH3),
39.1 (CH2), 60.7 (CH2), 106.9 (CH), 126.4 (C), 126.6 (CH), 134.7 (C),
160.4 (CO), 173.7 (CO). MS (ES+) m/z (%): 589 (8) [2M + Na]+,
306 (45) [M + Na]+, 284 (100) [M + H]+. Elemental analysis calcd
(%) for C12H17N3O5 (283.28): C, 50.88; H, 6.05; N, 14.83. Found: C,
51.02; H, 6.05; N, 15.10.
association to dimer only). Compounds 1-6 and 9 do not autoassociate
under the experimental conditions used.
Molecular Modeling. Molecular models for â-Xyl-Py-γ-Py-Ind and
r-Xyl-Py-γ-Py-Ind were generated employing the MM2* force field
as implemented in Macromodel 5.5. As a first step, a constrained
minimization was performed including the NOE derived distances as
experimental constrains. The obtained geometries were then subjected
to unconstrained minimizations, using 5000 conjugate gradients itera-
tions, and the GB/SA solvent model for the treatment of water.
DNA Interaction with ct-DNA of 1-3, 6, 7, and Netropsin. All
NMR interaction experiments have been carried out in phosphate buffer
(10 mM, pH ) 7). The binding abilities of the different ligands to calf
thymus DNA (ct-DNA) were investigated from 1H NMR titration
experiments. Ligand samples were prepared at a constant concentration
of 0.1 mM. The ct-DNA titrant sample (stock solution) was prepared
by dissolving 2 mg of ct-DNA into 0.7 mL solution of 0.1 mM in
ligand. The concentration of ct-ADN was calculated by UV/vis
EtO-γ-Py-Ind (23). A mixture of EtO-γ-Py-NO2 20 (10.0 g, 35.3
mmol) and 10% Pd-C (1.87 g, 5 mol %) was stirred in ethyl acetate
(200 mL) under an atmosphere of hydrogen at 45 psi for 24 h. The
catalyst was removed by filtration under gravity and the solvent
removed under reduced pressure to afford 9.90 g of brown oil. To a
solution of this oil in anhydrous CH2Cl2 (150 mL) was added 2-indole
carboxylic acid 22 (6.83 g, 42.4 mmol), DMAP (5.18 g, 42.4 mmol),
HOBt (5.72 g, 42.4 mmol), and DIPC (6.63 mL, 42.4 mmol) and stirring
under an argon atmosphere at room temperature overnight. The solvent
was removed under reduced pressure, and the residue purified by
column chromatography (SiO2, 50% EtOAc in hexane) to afford the
title compound 23 with a percentage of diisopropylurea (13.6 g, 23/
spectroscopy (c ) 3.58 × 10-3 M), using ꢀ260 ) 6600 M-1 cm-1 56
.
Then, 0.6 mL of the ligand solution was introduced into the NMR tube,
and increasing amounts of the titrant DNA solution were added. A
1D-NMR spectrum was recorded in the same “acquisition mode” (nt
) 256, Temp ) 26 °C) after each addition of ct-DNA. A progressive
broadening and disappearance of the proton signals from the ligand
was observed indicating the binding of the ligand to ct-DNA.
Blank Experiments. (1) A 0.1 mM solution in O-methyl-R-D-
glucopyranoside and 0.1 mM in sugar-oligoamide was used to prepare
a stock solution of ct-DNA under the same conditions as described
above. Then, a titration experiment was carried out. O-Methyl-R-D-
glucopyranoside resonances that appear between 3 and 4 ppm did not
experience any change in intensity or line widths, while oligoamide
resonances progressively disappeared upon addition of ct-DNA. (2) A
blank experiment was also carried out, titrating a 0.1 mM solution of
O-methyl-R-D-glucopyranoside with a stock solution of ct-DNA in the
same conditions. No changes in the glucopyranoside spectrum were
observed after addition of large amounts of DNA indicating that no
binding takes place.
1
urea (2.54:1) by H NMR by wt. 9.76 g, 70% yield of 23). The crude
material was used immediately for preparation of the acid 7 (see below).
To characterize a pure sample of the ester 23, the following reaction
was performed. A slurry of HO-γ-Py-Ind 7 (0.250 g, 0.679 mmol) in
absolute EtOH (10 mL) was treated with concentrated H2SO4 (3 drops)
and heated to reflux. Slow dissolution of the carboxylic acid substrate
was indicative of complete conversion (after 2 h), confirmed by TLC
analysis (EtOAc). The solvent was removed under reduced pressure,
and the residue was purified by column chromatography (SiO2, EtOAc)
to afford the title compound as crispy white foam (270 mg, 99%). An
analytical sample was prepared by trituration with EtOAc then Et2O
to give 23 as white waxy flakes (190 mg). Mp ) 165-166 °C; Rf )
0.596 (AcOEt 100%). IR (KBr): ν ) 1717, 1653, 1640 cm-1. 1H NMR
(200 MHz, DMSO): δ ) 1.18 (t, J ) 7.0 Hz, 3H), 1.75 (m, 2H), 2.33
(t, J ) 7.4 Hz, 2H), 3.21 (q, J ) 6.0 Hz, 2H), 3.83 (s, 3H), 4.05 (q, J
) 7.0 Hz, 2H), 6.88 (d, J ) 1.8 Hz, 1H), 7.05 (t, J ) 7.5 Hz, 1H),
7.19 (t, J ) 8.0 Hz), 7.29 (m, 2H), 7.46 (d, J ) 7.8 Hz, 1H), 7.65 (d,
J ) 7.8 Hz, 1H), 8.11 (t, J ) 5.8 Hz, 1H), 10.31 (s, 1H), 11.31 (s,
1H). 13C NMR (125 MHz, [D6] DMSO): δ ) 14.1 (CH3), 24.8 (CH2),
31.1 (CH2), 36.1 (CH3), 37.8 (CH2), 59.8 (CH2), 102.8 (CH), 104.1
(CH), 112.3 (CH), 118.2 (CH), 119.8 (CH), 121.6 (CH), 121.7 (C),
123.2 (C), 123.4 (CH), 127.2 (C), 131.7 (C), 136.6 (C), 158.3 (CO),
161.3 (CO), 172.8 (CO). MS (ES+) m/z (%): 816 (13) [2M + Na]+,
419 (4) [M + Na]+, 397 (100) [M + H]+. Elemental analysis calcd
(%) for C21H24N4O4 (396.4): C, 63.62; H, 6.10; N, 14.13. Found: C,
63.50; H, 6.10; N, 14.00.
Structural Studies. 1H NMR Experiments. All the spectra in
aqueous solution were recorded with presaturation of the water signal.
The chemical shifts were reported in ppm relative to DSS (0.00 ppm)
when D2O was used and relative to residual acetone (2.04 ppm) when
D2O/[D6]-acetone or H2O/[D6]-acetone was used in the experiment.
NMR structural studies of compounds 1-8 were based on monodi-
mensional and bidimensional (COSY, TOCSY, NOESY, ROESY)
spectra and were recorded at 300, 400, or 500 MHz and 26 °C in a
Varian instrument. Sample solutions were prepared at concentrations
ranging between 5 and 0.1 mM, depending on the solubility of the
compounds. The spectra for compound 7 were recorded at 0.1 mM in
D2O, where no self-association occurs.
Interaction Studies with poly(dA-dT) and poly(dG-dC). The
binding abilities of sugar-oligoamides 1 and 3 to poly(dA-dT) and
poly(dG-dC) were investigated by 1H NMR titration experiments.
Ligand samples were prepared at a constant concentration of 0.1 mM.
Poly(dA-dT) and poly(dG-dC) titrant samples (stock solution) were
prepared dissolving, in the case of poly (dA-dT), 50 unities of the
oligonucleotide in 2 mL of a solution 10-4 M of the ligand. Under
these conditions, the concentration of poly(dA-dT) was 3.79 mM bp
(ꢀ258 ) 13 200 M-1 cm-1).57 In the case of the Netropsin titration, the
concentration of poly(dA-dT) was 3.79 × 10-2 mM bp.
The concentration of poly (dG-dC) was 2.12 mM bp (ꢀ256 ) 16 800
M-1 cm-1 57 for all titrations, obtained by dissolving 25 unities of the
)
oligonucleotide in 0.5 mL of a 10-4 M solution of the ligand. Titrations
were carried out as described for ct-DNA.
Competition Experiment of 3 with Netropsin. Competition experi-
ments for 3 were carried out in D2O (phosphate buffer 10 mM, pH )
7) by 1H NMR titration in the Varian Inova-300 equipment. The
temperature was 26 °C. The free ligand sample was a saturated solution
of glyco-oligoamide in D2O, prepared by dissolving 1.5 mg of the ligand
in 1 mL. The concentration of the solution was 3 × 10-4 M. Then,
increasing amounts of a solution of ct-DNA 3.6 × 10-2 M were added
to the NMR tube until the complete disappearance of the resonances
Dilution Experiments. Dilution experiments of compounds 1-8
were carried out in D2O/[D6]-acetone, in H2O/D2O, and in D2O, at 26
°C. Concentrations of the sugar-oligoamides ranged between 10 mM
and 0.1 mM in D2O/[D6]-acetone. A dilution experiment in D2O started
with a saturated NMR sample of the sugar-oligoamide, due to its poor
solubility in water, and was diluted, until a concentration of 0.1 mM
was reached. Only compound 7 showed significant chemically induced
shift of the proton resonances upon dilution of the sample in D2O. Ka
) 100 M-1 was found for the autoassociation process of 7 (assuming
1
of 3 in the H NMR spectra was achieved. Then, increasing amounts
of one solution of Netropsin (10-2 M) were added.
TR-NOESY Experiments. TR-NOESY experiments for the bound
ligand were recorded on a 300 MHz spectrometer (Bruker) and 500
(56) Barton, J. K.; Goldberg, J. M.; Kumar, C. V.; Turro, N. J. J. Am. Chem.
Soc. 1986, 108, 2081.
(57) Lootiens, F. G.; Regenfuss, P.; Zechel, A.; Lieve, D.; Clegg, R. M.
Biochemistry 1990, 29, 9029-9039.
9
9532 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Carbohydrate-Based DNA Ligands
A R T I C L E S
MHz spectrometers (Varian and Bruker) with saturation of the residual
H2O signal or with the Watergate pulse sequence. Sugar oligoamide
solutions were prepared at saturation (1 mg/0.6 mL D2O), but the
operating concentration was 0.5 mM, after sonication of the NMR
sample. Further addition of the ct-DNA solution (2 mg/0.7 mL D2O or
H2O with 15% D2O) to the NMR sample induced broadening in the
resonance signals of the sugar-oligoamide. This fact is a clear
indication of binding of the ligand to ct-DNA.
CT98-0231), and RTN European project (HPRN-CT-2002-
00190). J.M. and C.S. thank the TMR project, and F.S. thanks
the RTN project for a postdoctoral fellowship. E.M. is grateful
to the Comunidad Autonoma de Madrid for a fellowship.
Supporting Information Available: Supporting Information
shows relevant NMR data of the free ligand, including inter-
and intrastrand NOEs, tables with chemical shifts, and details
on molecular modeling. Additionally, NMR titration of ligands
with ct-DNA and characterization of the bound state conforma-
tion by TR-NOESY experiments. This material is available free
of charge via the Internet at http://pubs.acs.org.
TR-NOESY experiments were recorded at 26 °C and performed with
mixing times of 200, 300, and 400 ms. The experiment in H2O/D2O
(15%) has been recorded at 18 °C.
Acknowledgment. Financial support for this work was
provided by the DGES (Grant BQU2000-1501-C01, BQU2003-
03550-C03-01 and -02), a European TMR project (FMRX-
JA050794N
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9533