Recognition of guanine-guanine mismatches by the dimeric form of 2-amino-1,8-naphthyridine

Article abstract of DOI:10.1021/ja0109186

Dimeric 2-amino-1,8-naphthyridine selectively binds to a G-G mismatch with high affinity (K_d = 53 nM). We have investigated a binding mechanism of naphthyridine dimer 2 to a G-G mismatch by spectroscopic studies, thermodynamic analysis, and str

Full text of DOI:10.1021/ja0109186

12650
J. Am. Chem. Soc. 2001, 123, 12650-12657
Recognition of Guanine-Guanine Mismatches by the Dimeric Form
of 2-Amino-1,8-naphthyridine
Kazuhiko Nakatani,*,† Shinsuke Sando,† Hiroyuki Kumasawa,† Jun Kikuchi,‡ and
Isao Saito*,†
Contribution from the Department of Synthetic Chemistry and Biological Chemistry,
Faculty of Engineering, Kyoto UniVersity, CREST, Japan Science and Technology Corporation (JST),
Kyoto 606-8501, Japan, and Protein Research Group, Genomic Sciences Center (GSC), RIKEN,
Yokohama Institute, Yokohama 230-0045, Japan
ReceiVed April 9, 2001
Abstract: Dimeric 2-amino-1,8-naphthyridine selectively binds to a G-G mismatch with high affinity (Kd )
53 nM). We have investigated a binding mechanism of naphthyridine dimer 2 to a G-G mismatch by
spectroscopic studies, thermodynamic analysis, and structure-activity studies for the thermal stabilization of
1
the mismatch. H NMR spectra of a complex of 2 with 9-mer duplex d(CATCGGATG)2 containing a G-G
mismatch showed that all hydrogens in two naphthyridine rings of 2 were observed upfield compared to those
of 2 in a free state. The 2D-NOESY experiments showed that each naphthyridine of 2 binds to a guanine in
the G-G mismatch within the π-stack. In CD spectra, a large conformational change of the G-G mismatch-
containing duplex was observed upon complex formation with 2. Isothermal calorimetry titration of 2 binding
to the G-G mismatch showed that the stoichiometry for the binding is about 1:1 and that the binding is
enthalpy-controlled. It is clarified by structure-activity studies that show (i) the linker connecting two
naphthyridine rings was essential for the stabilization of the G-G mismatch, (ii) the binding efficiency was
very sensitive to the linker structure, and (iii) the binding of two naphthyridines to each one of two Gs in the
G-G mismatch is essential for a strong stabilization. These results strongly supported the intercalation of
both naphthyridine rings of 2 into DNA base pairs and the formation of a hydrogen bonded complex with the
G-G mismatch.
Single nucleotide polymorphism (SNP) is a genetic difference
assays,10 rolling-circle amplifications,11 and MALDI-TOF mass
spectrometry12,13 have been developed for the detection of SNP.
To distinguish a mutant DNA from its wild type, these methods
for SNP detection require the sequence information, including
the site of a mutation. Because both mutant and wild type DNAs
are completely normal with respect to chemical structure and
properties, it is very difficult to detect a single nucleotide
difference of DNA sequences without knowing the sequence
information. There is a way to transform structurally normal
DNAs containing SNP into DNAs with structural defects. When
two DNAs containing a single nucleotide difference were mixed,
heat denatured, and annealed, a strand exchange between the
mutant and wild-type DNAs occurred to produce the hetero-
duplexes containing a base mismatch in addition to regeneration
of the original homoduplexes.14 Mismatch-containing DNAs are
considerably different from normal DNAs with respect to
structure,15-19 thermal stability,20-23 binding to mismatch repair
of a single nucleotide base. SNP widely spreads over a whole
genome sequence and exists in every 500-1000 base pairs.
Since a draft sequence of human genome was determined,1,2
SNP became extremely important as a genetic marker for the
identification of disease genes and detection of genetic muta-
tions.3,4 Thus, simple and rapid detection of a single nucleotide
difference in the DNA sequences is an indispensable technique
for both SNP mapping and typing. A number of high throughput
methods employing high-density arrays,5,6 single-primer exten-
sion assays,7 molecular beacons,8 TaqMan PCR,9 invader
* E-mail: nakatani@sbchem.kyoto-u.ac.jp.
† Faculty of Engineering, Kyoto University.
‡ Genomic Sciences Center, RIKEN.
(1) Lander, E. S. et al. Nature 2001, 409, 860-921.
(2) Venter, C. J. et al. Science 2001, 291, 1304-1351.
(3) Schafer, A. J.; Hawkins, J. R. Nat. Biotechnol. 1998, 16, 33-39.
(4) Collins, F. S.; Guyer, M. S.; Chakravarti, A. Science 1997, 278,
1580-1581.
(5) Wang, D. G.; Fan, J.-B.; Siao, C.-J.; Berno, A.; Young, P.; Sapolsky,
R.; Ghandour, G.; Perkins, N.; Winchester, E.; Spencer, J.; Kruglyak, L.;
Stein, L.; Hsie, L.; Topaloglou, T.; Hubbell, E.; Robinson, E.; Mittamann,
M.; Morris, M. S.; Shen, N.; Kilburn, D.; Rioux, J.; Nusbaum, C.; Roen,
S.; Hudson, T. J.; Lipshutz, R.; Chee, M.; Lander, E. S. Science 1998, 280,
1077-1082.
(10) Lyamichev, V.; Mast, A. L.; Hall, J. G.; James R.; Prudent, J. R.;
Kaiser, M. W.; Takova, T.; Robert, W.; Kwiatkowski, R. W.; Sander, T.
J.; Arruda, M.; Arco, D. A.; Neri, B. P.; Brow, M. A. D. Nat. Biotechnol.
1999, 17, 292-296.
(11) Lizardi, P. M.; Huang, X.; Zhu, Z.; Bray-Ward, P.; Thomas, D. C.;
Ward, D. C. Nat. Genet. 1998, 19, 225-232.
(6) For a review, see: Hacia, J. G. Nat. Genet. 1999, 21, 42-47
(supplement).
(12) Haff, L.; Smirnov, I. Genome Res. 1997, 7, 378-388.
(13) Ross, P.; Hall, L.; Smirnov, I.; Haff, L. Nat. Biotechnol. 1998, 16,
1347-1351.
(7) Shumaker, J. M.; Metspalu, A.; Caskey, C. T. Hum. Mutat. 1996, 7,
346-354.
(14) For a review of heteroduplex formation and analysis, see: Nataraj,
A. J.; Olivos-Glander, I.; Kusukawa, N.; Highsmith, W. E., Jr. Electro-
phoresis 1999, 20, 1177-1185.
(8) Thagi, S.; Fratu, D. P.; Kramer, F. R. Nat. Biotechnol. 1998, 16,
49-53.
(9) Morris, T.; Robertson, B.; Gallagher, M. J. Clin. Microbiol. 1996,
34, 2933-2936.
(15) Hunter, W. N.; Brown, T.; Kennard, O. J. Biomol. Struct. Dyn. 1986,
4, 173-191.
10.1021/ja0109186 CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/22/2001

Recognition of Guanine-Guanine Mismatches
J. Am. Chem. Soc., Vol. 123, No. 50, 2001 12651
enzymes,24,25 and susceptibility to drug interactions.26-32 These
physical and chemical properties characteristic of mismatches
are the keys for the detection of heteroduplexes.
Figure 1. Illustrations of proposed schemes for G bulge and G-G
mismatch recognition by 1 and its dimeric form, respectively. (a) G
bulge changes its conformation so as to bind 1 opposite the bulge. A
stacked-in conformation was shown as a representative of the initial
structures of a G bulge. (b) G-G mismatch binds to one naphthyridine
to produce a pseudo G bulge that allows the binding of another
naphthyridine.
(A), thymine (T), and cytosine (C) bulges, a formation of three
hydrogen bonds between 1 and G effectively discriminates a G
bulge from the others (Figure 1a). We have extended the
hypothesis to the recognition of a G-G mismatch by assuming
a G-G mismatched site as two consecutive G bulges. It has
been known that G-G mismatches form a stable base pair in
duplex DNA.35-37 However, we anticipated that an interaction
of 2-aminonaphthyridine to one of the Gs in the G-G mismatch
may produce a pseudo G bulge as a hypothetical intermediate
structure, which is subsequently bound by another molecule of
naphthyridine forming two hydrogen-bonded pairs of G-naph-
thyridine in a duplex π-stack (Figure 1b).
We have synthesized naphthyridine dimer 2 and investigate
its binding to a G-G mismatch in duplex DNA.38 DNase I
footprint titration showed that 2 selectively and strongly binds
to a G-G mismatch with the dissociation constant of 53 nM.
The affinity of 2 to G-A and G-T mismatches is about 360-
fold lower than that to a G-G mismatch. No apparent binding
of 2 to normal DNA containing only Watson-Crick base pairs
was observed. We have developed a mismatch-detecting sensor
useful for a surface plasmon resonance (SPR) assay by im-
mobilizing 2 onto the dextran-coated gold surface.38 With this
sensor, we have succeeded in differentiating 652 base pairs of
PCR products of a G/C heterozygote from those of a G/G
homozygote of HSP70-2 gene regarding the base at a nucleotide
number of 2345.39 For the accurate detection of SNP by
mismatch-binding drugs, recognition of each one of eight
mismatches by different drugs is ideal. Precise understanding
of the mechanism for the G-G mismatch recognition by 2
would provide useful information for the molecular design of
the drugs targeting other base mismatches. We herein report
We have reported that 2-amino-1,8-naphthyridine derivative
1 possessing the hydrogen bonding groups fully complementary
to guanine (G) binds selectively to a G bulge in duplex DNA.33,34
We have rationalized that 1binds to bulged G by forming three
hydrogen bonds and the hydrogen-bonded pair of 1-G is
stabilized in the duplex by stacking with the immediate base
pairs flanking the bulge. Because 1 binds to G but not to adenine
(16) Webster, G. D.; Sanderson, M. R.; Skelly, J. V.; Swann, P. F.; Li,
B. F.; Tickel, I. J. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 6693-6697.
(17) Gao, X.; Patel, D. J. J. Am. Chem. Soc. 1988, 110, 5178-5182.
(18) Nikonowicz, E. P.; Meadows, R. P.; Fagan, P.; Gorenstein, D. G.
Biochemistry 1991, 30, 1323-1334.
(19) Allawi, H. T.; SantaLucia, J., Jr. Biochemistry 1997, 36, 10581-
10594.
(20) Aboul-ela, F.; Koh, D.; Tinoco, I., Jr.; Martin, F. H. Nucleic Acids
Res. 1985, 13, 4811-4824.
(21) Li, Y.; Zon, G.; Wilson, W. D. Biochemistry 1991, 30, 7566-7572.
(22) Li. Y.; Agrawal, S. Biochemistry 1995, 34, 10056-10062.
(23) Allawi, H. T.; SantaLucia, J., Jr. Biochemistry 1998, 37, 2170-
2179.
(24) Fazakerley, G. V.; Qignard, E.; Woisard, A.; Guschlbauer, W.; van
der Marel, G. A.; van Boom, J. H.; Jones, M.; Radman, M. EMBO. J. 1986,
5, 3697-3703.
(25) Smith, J.; Modrich, P. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 4374-
4379.
(26) Kappen, L. S.; Goldberg, I. H. Proc. Natl. Acad. Sci. U.S.A. 1992,
89, 6706-6710.
(27) Kappen, L. S.; Goldberg, I. H. Biochemistry 1992, 31, 9081-9089.
(28) Chen, F.-U. Biochemistry 1998, 37, 3955-3964.
(29) Lee, S.-J.; Hurley, L. H. J. Am. Chem. Soc. 1999, 121, 8971-8977.
(30) Yang, X.-L.; Hubbard, R. B., IV; Lee, M.; Tao, Z.-F.; Sugiyama,
H.; Wang, A. H. Nucleic Acids Res. 1999, 27, 4183-4190.
(31) Jackson, B. A.; Alekseyev, V. Y.; Barton, J. K. Biochemistry 1999,
38, 4655-4662.
(35) Faibis, V.; Cognet, J. A.; Boulard, Y.; Sowers, L. C.; Fazakerley,
G. V. Biochemistry 1996, 35, 14452-14464.
(36) Lane, A. N.; Peck, B. Eur. J. Biochem. 1995, 230, 1073-1087.
(37) Borden, K. L.; Jenkins, T. C.; Skelly, J. V.; Brown, T.; Lane, A. T.
Biochemistry 1992, 31, 5411-5422.
(38) Nakatani, K.; Sando, S.; Saito, I. Nat. Biotechnol. 2001, 19, 51-
55.
(39) Milner, C. M.; Campbell, R. D. Immunogenetics 1990, 32, 242-
251.
(32) Jackson, B. A.; Barton, J. K. Biochemistry 2000, 39, 6176-6182.
(33) Nakatani, K.; Sando, S.; Saito, I. J. Am. Chem. Soc. 2000, 122,
2172-2177.
(34) Nakatani, K.; Sando, S.; Yoshida, K.; Saito, I. Bioorg. Med. Chem.
Lett. 2001, 11, 335-337.

12652 J. Am. Chem. Soc., Vol. 123, No. 50, 2001
Nakatani et al.
Figure 3. 2D-NOESY spectrum for the aromatic region of the 2-GG1
complex.
Figure 2. 2D-TOCSY spectrum for the aromatic region of the 2-GG1
complex.
Table 1. Chemical Shifts (ppm) of Aromatic Hydrogens of
Naphthyridine in 2 and 2-GG1 Complex
spectroscopic and thermodynamic analysis for the binding of 2
to G-G mismatch-containing DNAs. All the data described here
showed that the cooperative binding of two naphthyridines to a
G-G mismatch through the hydrogen bonding within the DNA
π-stack is the basis for the formation of stable complexes
between the drugs and mismatches.
H3
H4
H5
H6
7-Me
2a
8.28
8.16
7.91
8.11
7.77
7.54
8.09
7.27
7.32
7.33
6.53
6.73
2.67
2.37
1.75
2-GG1b,c
a In CD3OD. b In D2O. c Two sets of signals were observed.
Results and Discussion
Structural Analysis of the 2-G-G Mismatch Complex
by NMR. To know the mode of the binding of naphthyridine
dimer to a G-G mismatch, we have measured 1H and 31P NMR
of the complex produced from 2 and self-complementary 9-mer
duplex d(CATCGGATG)2 (GG1) containing a G-G mismatch
in the middle of the sequence.36 The central domain of d(CGG)/
(CGG) in the sequence is the site with the highest affinity for
2 binding as we previously determined by DNase I footprint
titration.38 We first focused our attention on the aromatic
hydrogens of 2 in the complex. The 2D-TOCSY spectrum for
the aromatic regions of the 2-GG1 complex was shown in
Figure 2. Because two naphthyridine rings of 2 are symmetric
to each other, four aromatic hydrogens corresponding to H3,
H4, H5, and H6 were observed in CD3OD at 8.28, 8.11, 8.09,
and 7.33 ppm, respectively. In contrast, the aromatic hydrogens
of two naphthyridines in the 2-GG1 complex were separately
observed, showing that two naphthyridines are no longer
symmetric to each other in the complex. Thus, two sets of cross-
peaks for H5-H6 and H3-H4 were observed in the 2D-TOCSY
spectrum of the 2-GG1 complex. Although signal intensities
of H3-H4 cross-peaks were rather weak compared to those
for H5-H6, distinct cross-peaks of H3-H4 in addition to those
of H4-H5 and H5-H6 were observed in the 2D-NOESY
spectrum (Figure 3). Each H6 signal of two naphthyridines in
the 2-GG1 complex observed at 6.53 and 6.73 ppm was
unambiguously assigned on the basis of a correlation with C7-
methyl hydrogens observed at 2.37 and 1.75 ppm, respectively.
The chemical shifts of all the eight hydrogens of two naphthy-
ridines and C7-methyl hydrogens were summarized in Table 1.
As clearly seen from the table, all signals showed substantial
upfield shifts in the 2-GG1 complex compared to those in a
free state of 2. The upfield shifts of naphthyridine hydrogens
are attributable to the ring-current effects from the aromatic
rings, suggesting that both naphthyridine rings of 2 intercalated
into DNA base pairs in the complex. The formation of hydrogen
bonds between 2 and GG1 is also suggested by a marked change
Figure 4. Conventional numbering of the nucleotide bases in the
2-GG1 complex.
1
of exchangeable protons in H NMR in terms of the chemical
shifts and the thermal profile of the signals (Figure S1).
To identify the intermolecular interactions, the exchangeable
protons involved in the hydrogen bonding of 2 and GG1 were
identified using 2D-NOESY experiments. A conventional
numbering of the nucleotide bases in the 2-GG1 complex for
the NMR assignments was shown in Figure 4. The chemical
shifts of the imino protons for G1, T2, T3, and G4 in a free
state of GG1 were reported as 12.8, 13.7, 13.6, and 12.6 ppm,
respectively.36 The imino protons of T2 (T2NH), T3 (T3NH),
and G4 (G4NH) in the 2-GG1 complex were assigned at 13.4,
12.6, and 12.0 ppm, respectively, by observing distinct cross-
peaks of T2NH with T3NH and T3NH with G4NH (Figure 5c).
The signal at 12.0 ppm showed a correlation with two different
imino protons at 11.75 and 11.65 ppm. A correlation between
the signal at 11.65 ppm and T7NH, which is identical to T3NH
in the chemical shift, identifies the imino proton as G6NH. The
signal at 11.75 ppm is likely an imino proton of one of the
guanines (G5 or G5′) bound to naphthyridine. These assignments
of the imino protons were supported by the correlations between
the imino protons of T2, T3, and G4 and an aromatic hydrogen
at the C2 position of adenines A2 (A2CH) and A3 (A3CH)
(Figure 5b). Thus, A2CH showed strong correlations with T2NH
and T3NH, whereas A3CH correlated to all the imino protons
of T2, T3, and G4. Because of fast exchange with water protons
at both terminuses (G1NH and G9NH), these imino protons were
not observed at 17 °C. Similarly, signal intensities for neighbor-
ing imino protons at T2NH (T8NH is overlapped) decreased
with the temperature change over 37 °C (Figure S1).
A strong correlation between the signals at 12.0 and 9.00
ppm was particularly suggestive for the signal assignment,
because the proton at 9.00 ppm had a correlation to the methyl

Recognition of Guanine-Guanine Mismatches
J. Am. Chem. Soc., Vol. 123, No. 50, 2001 12653
(G5NH) and G5′ (G5′NH), respectively. The imino proton of
G5′ was overlapped by the signal of G4NH. Accordingly, the
methyl group observed at 2.37 ppm was attached to the
naphthyridine bound to G5 (ND5), whereas the methyl group
of the other naphthyridine bound to G5′ (ND5′) appeared at
1.75 ppm. A very significant cross-peak between the methyl
hydrogen of ND5 (ND5CH3) and the amino hydrogen of G5′
(G5′NH2) suggests a stacking of the hydrogen-bonded pairs of
ND5-G5 and G5′-ND5.
The remaining signal at 10.35 ppm was assigned as the two
amide protons of 2 (ND5NH and ND5′NH) bound to the
carbonyl oxygens of G5 and G5′, on the basis of the correlations
to the methylene hydrogens in the linker connecting two
naphthyridines. Characteristic correlations to NDNH are ob-
served for the signals at 7.53 and 7.90 ppm, which were assigned
as the amino protons of C6 (C6NH2) and C4 (C4NH2),
respectively. These assignments were supported by the correla-
tions of G5NH-C4NH2, G4NH-C4NH2, and G6NH-C6NH2
(Figure 5b). The correlations of ND5NH to G5NH and ND5′NH
to G5′NH were also observed, although the two correlations
were very much different in their intensities. NOE correlations
observed for the 2-GG1 complex are consistent with the
proposed structure of the complex, where each one of two
naphthyridines binds to a guanine in the G-G mismatch and
the naphthyridine-guanine pairs are stacked with each other.
The observed NOEs were illustrated on the proposed chemical
structure of the 2-GG1 complex (Figure 6). A marked highfield
shift of ND5′CH3 (1.75 ppm) compared to the shift of ND5CH3
(2.37 ppm) suggests that ND5CH3 extrudes into a minor groove,
whereas ND5′CH3 is probably stacked with the neighboring
G-ND5 and/or G-C base pair and more strongly affected by
the ring-current shielding effects. This asymmetry regarding the
position of two naphthyridines in the complex reflects on an
appearance of the NOE cross-peak for ND5CH3-G5′NH2 and
the disappearance of a ND5′CH3-G5NH2 cross-peak. The
tetrahedral structure of the basic amino nitrogen of the linker
and insufficient linker length between two naphthyridines are
highly likely the reasons for the asymmetric location of two
naphthyridine-guanine pairs in the 2-GG1 complex.
31P NMR spectra of GG1 and the 2-GG1 complex were
measured at 17 and 37 °C (Figure 7). The broad signals of GG1
were observed between -2.7 and -2.2 ppm at 37 °C. The
narrow distribution of 31P chemical shifts (i.e., 0.5 ppm)
indicates that dihedral angles of phosphate backbones are those
in B- and A-type DNA structures at 37 °C. As the temperature
decreased to 17 °C, the signals became more discrete and sharp
in shape. This observation correlates well to the melting of a
duplex form of GG1 to a single strand, because the melting
temperature (Tm) of GG1 was reported to be 35 °C.36 In contrast,
the 31P signals of the 2-GG1 complex shifted upfield by ∼1
ppm compared to those of free GG1 and were less sensitive to
a temperature change because of the increased Tm of the duplex
upon complex formation. Because 31P chemical shifts in nucleic
acids provide a probe of the conformation of a phosphate
ester,18,40 wider distribution of 31P signals in the complex
compared to its free state may imply conformational changes
of a phosphate backbone upon complex formation.
Figure 5. 2D-NOESY spectrum of the 2-GG1 complex involving
the exchangeable protons.
hydrogens at 1.75 ppm in one of two naphthyridines (Figure
5a). The other methyl hydrogens were observed at 2.37 ppm
and strongly correlated to the signal at 9.95 ppm. On the basis
of a putative hydrogen-bonded guanine and 2-aminonaphthy-
ridine pair, the signals observed at 9.95 and 9.00 ppm were
assigned as the hydrogens of the amino group at the C2 position
of the guanines (G5NH2 and G5′NH2) bound to naphthyridine.
The two cross-peaks with a strong intensity were assigned to
the correlations between an imino proton and an amino proton
of the guanine bound to naphthyridine. Thus, two signals at
11.75 and 12.0 ppm were assigned as imino protons of G5
Conformational Change of a G-G Mismatch-Containing
Duplex upon 2 Binding. The NMR spectra of the 2-GG1
complex showed that both naphthyridine rings intercalated in
DNA base pairs. These observations implied that the structure
of a mismatched site in GG1 would be significantly changed
(40) Roongta, V. A.; Jones, C. R.; Gorenstein, D. G. Biochemistry 1990,
29, 5245-5258.

12654 J. Am. Chem. Soc., Vol. 123, No. 50, 2001
Nakatani et al.
Figure 8. CD spectra of DNA duplexes d(CTA ACG GAA TG)/
d(CAT TCX GTT AG) containing G-G (GG2: X ) G) and G-A
(GA1: X ) A) mismatches in the absence and presence of 2 (9.1 µM).
CD spectra of duplexes (100 µM base concentration) were measured
in 10 mM sodium cacodylate buffer (pH 7.0) and 100 mM NaCl at 25
°C.
shifted to 230 nm in the 2-GG2 complex. Intense induced CD
was also observed in the region from 300 to 360 nm, showing
that naphthyridine rings are located within a chiral environment
produced by a DNA double helix. Similar induced CD bands
but with much weaker intensity were detected for the 2-GA1
complex. We have reported that the binding of 2 to the G-G
mismatch in the d(CGG)/d(CGG) is about 360-fold stronger than
that to the G-A mismatch in d(CGG)/d(CAG).38 The spectral
change induced by the addition of 2 is much more significant
for GG2 than for GA1, showing a good correlation between
the CD spectral change and the efficiency of 2 binding to the
mismatched site. These experiments clearly showed that binding
of 2 to a G-G mismatch accompanied a significant conforma-
tional change.
Figure 6. Illustration of chemical shifts of protons and NOEs observed
for the 2-GG1 complex.
Thermodynamics for the Formation of the 2-GG2
Complex. We have analyzed energetics and stoichiometry of
the interaction between 2 and GG2 by isothermal titration
calorimetry (ITC).41,42 In the experiments, the calorimeter cell
was filled with the solution of GG2, and the solution of 2 was
injected. Evolution of heat upon injection of 2 was measured
as the power (µcal s-1) that is required to maintain a zero
temperature difference between the sample and the reference
cell. Each injection of 2 into the solution containing GG2
produced a large exothermic heat of reaction, whereas the
dilution heat produced by adding 2 into buffer was small and
negligible (Figure 9). Molar heat values obtained by integration
of the data were plotted as a function of the molar ratio between
2 and GG2 ([2]/[GG2]) (Figure 9, inset). From the best fit for
the binding isotherm using a single site model, thermodynamic
parameters were obtained. An apparent association constant (Ka)
for the binding of 2 to GG2 was 0.9 × 107 M-1, which is
consistent with Ka of 1.9 × 107 M-1 (Kd of 53 nM) obtained
by DNase I footprint titration within experimental variability.38
The stoichiometry (n) for the binding was 1.2, suggesting a 1:1
binding. The free energy change at 25 °C by the formation of
the 2-GG2 complex was -9.5 ( 0.3 kcal/mol, which consisted
of an enthalpy gain of -36.5 ( 0.4 kcal/mol and an entropy
Figure 7. 31P NMR spectra of (a) GG1 and (b) 2-GG1 complex at
17 and 37 °C. Signals shown with an asterisk are the internal standard.
upon complex formation. CD spectra of short 11-mer duplexes
d(CTA ACG GAA TG)/d(CAT TCX GTT AG) containing
G-G (GG2: X ) G) and G-A (GA1: X ) A) mismatches
were measured in the presence and absence of 2 (Figure 8).
While CD spectra of GG2 are quite similar to those of GA1
and the fully matched duplex (X ) C, data not shown), addition
of 2 mol equiv of 2 into GG2 induced a dramatic change of
the spectra. The positive maximum at 270 nm increased the
intensity by 2-fold, and negative bands at 250 nm of GG2
(41) Haq, I.; Trent, J. O.; Chowdhry, B. Z.; Jenkins, T. C. J. Am. Chem.
Soc. 1999, 121, 1768-1779.
(42) Ren, J.; Jenkins, T. C.; Chaires, J. B. Biochemistry 2000, 39, 8439-
8447.

Recognition of Guanine-Guanine Mismatches
J. Am. Chem. Soc., Vol. 123, No. 50, 2001 12655
Scheme 1a
Figure 9. Isothermal titration carlorimetry (ITC) for the binding of 2
to GG2 at 25 °C. Titration was conducted by adding 10 µL of 2 (70
µM) every 3 min into a buffer solution (10 mM sodium cacodylate pH
7.0, 100 mM NaCl) containing G-G mismatch DNA GG2 (4 µM)
(blue line) or buffer alone (black line). Inset: Molar heat values plotted
as a function of the [2]/[GG2] molar ratio. The solid red line represents
the best fit binding isotherm. The data were fitted using a single site
model.
a Reagents: (a) Boc2O, CHCl3, 83%; (b) (i) aq NaOH, THF; (ii)
N-hydroxysuccinimide, EDCI, DMF, 85%; (c) 2-amino-7-methyl-1,8-
naphthyridine, CHCl3, 46%; (d) HCl, AcOEt, CHCl3, quantitative; (e)
pentafluorophenyl N-Boc-3-aminopropanoate, diisopropylethylamine,
DMF, 77%.
Scheme 2a
loss of 27.0 ( 0.7 kcal/mol. These data showed enthalpy-driven
binding of 2 to a G-G mismatch.
Synthesis of Probe Molecules 3 and 4. Having confirmed
the intercalation of two naphthyridines of 2 into the DNA
π-stack and the hydrogen bonding between naphthyridine and
guanine, we then examined the effect of the structural modifica-
tion of drugs on the efficiency of the binding by using two probe
molecules, 3 and 4. In naphthyridine dimer 3, the secondary
amino group in the linker of 2 connecting two naphthyridine
rings was converted to a tertiary amide. To keep the magnitude
of the positive charge the same between 2 and 3, a primary
amino group was additionally incorporated into the linker. The
other probe molecule 4 is the 2-aminoquinoline-naphthyridine
hybrid, where one of two 2-aminonaphthyridines in 2 was
replaced with 2-aminoquinoline. Synthetic schemes for 2, 3,
and 4 were shown in Scheme 1. Methanolysis of commercially
available 3-[(2-cyanoethyl)amino]propanenitrile gave dimethyl
ester 5. Reaction of 5 with di-tert-butyl dicarbonate afforded 6
in 83% yield. Hydrolysis of 6 with aqueous sodium hydroxide
followed by the reaction with N-hydroxysuccinimide in the
presence of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride (EDCI) gave activated diester 7 in 85%. Con-
densation of 7 with 2-amino-7-methyl-1,8-naphthyridine af-
forded Boc-protected dimeric naphthyridine 8 in 46%, which
was hydrolyzed for deprotection to give dimeric naphthyridine
2. Coupling of 2 with pentafluorophenyl N-Boc-3-aminopro-
panoate followed by acid treatment produced 3 in 77%.
Bispentafluorophenoxy ester 10 was used instead of 7 for the
synthesis of naphthyridine-aminoquinoline hybrid 4 (Scheme
2). Conversion of 5 to 10 was achieved in three steps in 57%
overall yield. Subsequent coupling of 10 with 1 equiv of
2-amino-7-methylnaphthyridine produced monopentafluoroester
11 in 90%. A second coupling of 11 with 2-aminoquinoline
proceeded rather sluggishly to give protected naphthyridine-
aminoquinoline hybrid 12 in 34% yield, which was converted
to 4 under acidic conditions.
a Reagents: (a) (i) Boc2O, CHCl3; (ii) aq NaOH, THF; (ii) pen-
tafluorophenol, EDCI, DMF, 57%; (b) 2-amino-7-methyl-1,8-naphthy-
ridine, DMF, 90%; (c) 2-aminoquinoline, DMF, 34%; (d) HCl, AcOEt,
CHCl3, quantitative.
mismatch was examined by the difference in Tm (∆Tm) between
the presence and absence of drugs (Table 2). ∆Tm increases as
the affinity of drug to a G-G mismatch increases. All drugs
1-4 did not stabilize fully matched duplex GC1, whereas the
degree of the stabilization of GG2 was varied between the drugs.
In marked contrast to a large ∆Tm of 16.4 °C upon 2 binding to
GG2, naphthyridine monomer 1 only weakly stabilized GG2,
showing a ∆Tm of 1.8 °C under the same conditions. Duplex
stabilization by dimeric naphthyridine was highly sensitive to
the structure of the linker by comparing ∆Tm values obtained
in the presence of 2 and 3. A transformation of a secondary
amine in the linker of 2 to a tertiary amide in 3 resulted in a
decrease of ∆Tm by 6.2 °C. Acylation of the amines changes
Effects of the Drug Structure on the Mismatch Binding.
The efficiency of the drug binding to GG2 containing a G-G

12656 J. Am. Chem. Soc., Vol. 123, No. 50, 2001
Nakatani et al.
Table 2. Difference of Melting Temperature (∆Tm) of G-G
Mismatch-Containing Duplexes in the Presence and Absence of
Druga,b
for the stabilization of the mismatch by 2 as we reported
earlier.28 The sequence dependent stabilization of the G-G
mismatch by 2 suggests that the stacking of the naphthyridine-
guanine pair by the flanking base pairs plays an important role
for the stabilization of the complex. For the molecular design
of the drugs targeting G-A and G-T mismatches, a hybrid
drug consisting of naphthyridine and a planar molecule with
hydrogen bonding groups fully complementary to A and T
would be potentially attractive.
drugc
5′-CTAACGGAATG-3′
3′-GATTGXCTTAC-5′
1
2
3
4
GG2: X ) G
GC1: X ) C
1.8
-0.3
16.4
-0.3
10.2
-0.2
2.4
-0.7
a The UV melting curve was measured at a total base concentration
of 100 µM in 10 mM sodium cacodylate buffer (pH 7.0) containing
0.1 M NaCl. ∆Tm is the difference in Tm of the duplex obtained in the
presence and absence of drug. Tm values of G-G mismatched and fully
matched duplexes were 25.8 and 42.5 °C, respectively. b Mismatched
bases are shown in boldface. c Concentration of drug 1 was 18.2 µM,
whereas those of 2, 3, and 4 were 9.1 µM.
Experimental Section
Methyl 3-((2-(Methoxycarbonyl)ethyl)amino)propanoate (5). A
solution of acetyl chloride (14.9 g, 0.19 mol) in MeOH (70 mL) was
heated under reflux for 5 min. To the solution was added 3,3′-iminobis-
(propionitrile) (5.0 g, 0.04 mol), and the mixture was heated under
reflux for 8 h. Additional acetyl chloride (5.5 g, 0.07 mol) was added,
and the mixture was heated under reflux for an additional 3 h. The
reaction mixture was left at room temperature overnight and filtered,
and the filtrate was evaporated in vacuo to give hydrochloride of 5
(8.1 g, 88%) as a white granular solid: 1H NMR (D2O, 400 MHz) δ
the structure from tetrahedral to trigonal planar in amide, where
three atoms directly attached to the nitrogen are located on the
same plane. Therefore, a weaker stabilization of GG2 by 3 is
most likely due to the increase of the energy necessary for a
conformational change from a free to bound state of 3 compared
to the energy required for 2. Only a small ∆Tm of 2.4 °C was
observed for GG2 in the presence of naphthyridine-amino-
quinoline hybrid 4. We previously observed almost negligible
stabilization of DNA containing a G-bulge by a 2-aminoquino-
line derivative.33 This was rationalized by a steric repulsion
between C8-H of 2-aminoquinoline and N2-H of G in a
postulated 2-aminoquinoline-G hydrogen bonded pair. The
stabilization of GG2 was much stronger for 2 and 3, containing
two naphthyridine moieties in one molecule, than 4. These
observations strongly suggested that the binding of two naph-
thyridines to each G in the G-G mismatch is essential for the
strong stabilization of the mismatch. Only a weak stabilization
of GG2 by 1 emphasized a significance of the covalent
connection of two naphthyridines. The importance of three
hydrogen bonds between 2 and guanine in the 2-GG1 complex
for the strong duplex stabilization is also suggested by using
oligomers containing inosine (I)-inosine and guanine-inosine
mismatches. Melting temperatures of d(CTA ACX GAA TG)/
d(CAT TCI GTT AG) containing a I-I mismatch (X ) I) and
a G-I mismatch (X ) G) increased only 10.4 and 12.8 °C,
respectively, in the presence of 2, substantially lower than the
increase of 16.4 °C for the G-G mismatch. Although a
replacement of guanine by inosine affects not only the hydrogen
bonding but also the stacking interactions, it is estimated that
three hydrogen bonds in each naphthyridine-guanine pair
strongly contribute to the stabilization of a G-G mismatch-
containing duplex.
Implication for the Design of Mismatch Binding Mol-
ecules. The binding of dimeric naphthyridine 2 to G-G
mismatch was examined by NMR and CD spectra with respect
to the complex structure, and by ITC regarding the thermo-
dynamics. All the data obtained by these analyses showed that
2 and G-G mismatches form a stable complex with 1:1
stoichiometry, where each one of two naphthyridines of 2 binds
to a guanine and the two naphthyridine-guanine pairs are
stacked with each other in the duplex. The structure of the linker
connecting two naphthyridines was very sensitive to the binding,
as clearly shown by the comparison of ∆Tm obtained for 2 and
3. Furthermore, a loss of full complementarity of the hydrogen
bonding groups to those of a mismatched base considerably
decreases the affinity to the mismatch. The marked difference
between 1 and 2 regarding the efficiency for the binding to a
G-G mismatch strongly suggested that a cooperative binding
of two naphthyridines to a G-G mismatch is essential for the
formation of the stable complex. In addition to these factors,
the sequence flanking the G-G mismatch has significant effects
3.60 (s, 6 H), 3.24 (t, 4 H, J ) 6.4 Hz), 2.73 (t, 4 H, J ) 6.4 Hz); 13
C
NMR (D2O, 100 MHz) δ 172.9, 52.7, 43.1, 30.0; FABMS (NBA), m/e
189 [(M + H)+]; HR-FABMS calcd for C8H15O4N [(M + H)+],
189.1000; found, 189.1006.
Methyl 3-((tert-Butoxy)-N-(2-(methoxycarbonyl)ethyl)carbonyl-
amino)propanoate(6). To a suspension of the hydrochloride of 5 (1.0
g, 4.4 mmol) in CHCl3 (30 mL) was added triethylamine (0.73 g, 7.2
mmol) at 0 °C. After stirring for 1 h, the reaction mixture was filtered,
and the filtrate was evaporated in vacuo to give free 3,3′-iminobis-
(methyl propionate) as a white solid. To the solution of this material
in dry CHCl3 (6 mL) was added di-tert-butyl dicarbonate (1.06 g, 4.9
mmol) slowly at 0 °C, and the mixture was stirred overnight at room
temperature. The solution was washed successively with saturated
aqueous KHSO4, saturated aqueous NaHCO3, and saturated aqueous
NaCl and dried over MgSO4. Solvent was evaporated to give 6 (1.06
g, 83%) as an oil: 1H NMR (CDCl3, 400 MHz) δ 3.64 (s, 6 H), 3.46
(m, 4 H), 2.53 (m, 4 H), 1.41 (s, 9 H); 13C NMR (CDCl3, 100 MHz,
50 °C) δ 172.1, 154.9, 79.9, 51.5, 44.0, 33.5, 28.2; EIMS, m/e (%)
289 (M+) (3), 233 (20), 202 (45), 174 (25), 57 (100); HRMS calcd for
C13H23NO6 (M+), 289.1525; found, 289.1515.
2,5-Dioxopyrrolidinyl 3-((tert-Butoxy)-N-(2-((2,5-dioxopyrrolidi-
nyl)oxycarbonyl)ethyl)carbonylamino)propanoate (7). To a stirred
solution of 6 (283 mg, 0.98 mmol) in THF (1.25 mL) was added 2 N
aqueous NaOH solution (1.25 mL, 3 mmol), and the mixture was
vigorously stirred at room temperature for 20 h. MeOH (1.25 mL) was
added, the reaction mixture was neutralized to pH 7 with Amberlite
IR-120 (H+ form) ion-exchange resin, and the whole mixture was stirred
overnight. The mixture was readjusted to pH 3.5 with the resin and
filtered, and the solvent was evaporated to give N-(tert-butoxycarbonyl)-
imino-3,3′-bis(propionic acid) as a white solid. To the solution of acid
in dry DMF (5 mL) were added N-hydroxysuccinimide (282 mg, 2.5
mmol) and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydro-
chloride (470 mg, 2.5 mmol). The mixture was stirred for 48 h at room
temperature. The solvent was evaporated, and the crude product was
suspended in CHCl3. The organic layer was washed successively with
H2O, saturated aqueous NaHCO3, and saturated aqueous NH4Cl and
dried over MgSO4. The solvent was evaporated to give 7 (358 mg,
85%) as a white solid: 1H NMR (CDCl3, 400 MHz) δ 3.59 (m, 4 H),
2.89 (m, 2 H), 2.83 (m, 2 H), 2.79 (s, 8 H), 1.43 (s, 9 H); 13C NMR
(CDCl3, 100 MHz) δ 168.9, 167.3, 154.8, 80.9, 44.6, 36.5, 30.8, 28.3,
25.6; FABMS (NBA), m/e 456 [(M + H)+]; HR-FABMS calcd for
C19H26O10N3 [(M + H)+], 456.1616; found, 456.1591.
3-((tert-Butoxy)-N-(2-(N-(7-methylpyridino[3,2-e]pyridin-2-yl)-
carbamoyl)ethyl)carbonylamino)-N-(7-methylpyridino[3,2-e]pyri-
din-2-yl)propanoate (8). To a solution of 7 (313 mg, 0.74 mmol) in
dry CHCl3 (30 mL) was added 2-amino-7-methyl-1,8-naphthyridine
(294 mg, 1.9 mmol), and the mixture was stirred at room temperature
for 48 h. The mixture was filtered, the filtrate was evaporated to dryness,
and the crude residue was purified by silica gel column chromatography
to give 8 (185 mg, 46%) as a pale yellow solid: 1H NMR (CDCl3, 400
MHz) δ 8.78 (br, 2 H), 8.45 (d, 2 H, J ) 8.8 Hz), 8.13 (d, 2 H, J )

Recognition of Guanine-Guanine Mismatches
J. Am. Chem. Soc., Vol. 123, No. 50, 2001 12657
8.8 Hz), 7.98 (d, 2 H, J ) 8.4 Hz), 7.25 (d, 2 H, J ) 8.4 Hz), 3.65 (t,
4 H, J ) 6.8 Hz), 2.77 (t, 4 H), 2.74 (s, 6 H), 1.37 (s, 9 H); 13C NMR
(CDCl3, 100 MHz) δ 171.2, 163.2, 155.4, 154.4, 153.9, 139.4, 121.8,
118.8, 115.0, 80.5, 44.8, 37.7, 28.8, 26.0; FABMS (NBA), m/e 544
[(M + H)+]; HR-FABMS calcd for C29H34O4N7 [(M + H)+], 544.2645;
found, 544.2670.
To a solution of 9 (1.5 g, 2.5 mmol) in dry DMF (5 mL) was added
2-amino-7-methyl-1,8-naphthyridine (180 mg, 1.1 mmol) and diiso-
propylethylamine (163 mg, 1.26 mmol). The mixture was stirred at
room temperature for 15 h. The solvent was evaporated to dryness,
and the crude residue was purified by silica gel column chromatography
to give 11 (1.29 g, 90%) as a pale yellow solid: 1H NMR (CDCl3, 400
MHz) δ 9.01 (br, 1 H), 8.44 (d, 1 H, J ) 8.8 Hz), 8.12 (d, 1 H, J )
8.8 Hz), 7.99 (d, 1 H, J ) 8.0 Hz), 7.26 (d, 1 H, J ) 8.0 Hz), 3.66 (m,
4 H), 2.90 (m, 2 H), 2.74 (m, 2 H), 2.73 (s, 3 H), 1.42 (s, 9 H); FABMS
(NBA), m/e 569 [(M + H)+]; HR-FABMS calcd for C26H26O5N4F5
[(M + H)+], 569.1821; found, 569.1827. To a solution of 11 (587 mg,
1.0 mmol) in dry DMF (2 mL) was added 2-aminoquinoline (180 mg,
1.2 mmol) and diisopropylethylamine (160 mg, 1.2 mmol). The mixture
was stirred at room temperature for 48 h. The solvent was evaporated
to dryness, and the crude residue was purified by silica gel column
chromatography to give 12 (187 mg, 34%) as a pale yellow solid: 1H
NMR (CDCl3, 400 MHz) δ 8.95 (s, 1 H), 8.73 (s, 1 H), 8.44 (d, 1 H,
J ) 8.8 Hz), 8.27 (d, 1 H, J ) 8.0 Hz), 8.12 (d, 1 H, J ) 8.8 Hz), 8.07
(d, 1 H, J ) 8.8 Hz), 7.95 (d, 1 H, J ) 8.0 Hz), 7.75 (d, 1 H, J ) 8.8
Hz), 7.71 (d, 1 H, J ) 8.0 Hz), 7.61 (t, 1 H, J ) 7.0 Hz), 7.40 (t, 1 H,
J ) 7.0 Hz), 7.19 (d, 1 H, J ) 8.0 Hz), 3.64 (m, 2 H), 3.63 (m, 2 H),
2.74 (m, 2 H), 2.73 (m, 2 H), 2.72 (s, 3 H), 1.38 (s, 9 H); 13C NMR
(CDCl3, 100 MHz) δ 170.6, 170.4, 163.2, 155.4, 154.4, 153.3, 150.8,
146.5, 139.1, 138.6, 136.5, 129.9, 127.5, 127.3, 126.9, 125.1, 121.6,
118.6, 114.4, 114.3, 80.4, 77.2, 44.7, 37.3, 30.9, 28.3, 25.5; FABMS
(NBA), m/e 529 [(M + H)+]; HR-FABMS calcd for C29H33O4N6 [(M
+ H)+], 529.2561; found, 529.2565.
N-(7-Methylpyridino[3,2-e]pyridin-2-yl)-3-((2-(N-(2-quinolyl)car-
bamoyl)ethyl)amino)propanamide (4). To a solution of 12 (116 mg,
0.22 mmol) in CHCl3 (3 mL) was added ethyl acetate containing 4 M
HCl (2 mL) at 0 °C, and the mixture was stirred at room temperature
for 0.5 h. The solvent was evaporated to dryness to give hydrochloride
of 4 (quantitative yield) as a white solid: 1H NMR (CD3OD, 400 MHz)
δ 8.94 (d, 1 H, J ) 8.4 Hz), 8.92 (d, 1 H, J ) 8.4 Hz), 8.70 (d, 1 H,
J ) 8.8 Hz), 8.65 (d, 1 H, J ) 8.8 Hz), 8.21 (d, 1 H, J ) 8.4 Hz), 8.18
(d, 1 H, J ) 8.4 Hz), 8.05 (t, 1 H, J ) 7.2 Hz), 7.85 (d, 1 H, J ) 8.4
Hz), 7.83 (t, 1 H, J ) 7.2 Hz), 7.65 (d, 1 H, J ) 8.4 Hz), 3.60 (t, 2 H,
J ) 6.4 Hz), 3.55 (t, 2 H, J ) 6.4 Hz), 3.31 (t, 2 H, J ) 6.4 Hz), 3.23
(t, 2 H, J ) 6.4 Hz), 2.99 (s, 3 H); 13C NMR (CD3OD, 100 MHz) δ
174.5, 172.4, 161.4, 158.3, 150.6, 148.6, 147.5, 145.2, 141.5, 135.4,
134.2, 130.2, 129.5, 126.5, 123.7, 121.5, 118.8, 118.2, 114.9, 44.6,
44.0, 33.9, 33.8, 20.8; FABMS (NBA), m/e 429 [(M + H)+]; HR-
FABMS calcd for C24H25O2N6 [(M + H)+], 429.2037; found, 429.2038.
Isothermal Titration Calorimetry (ITC). ITC titration experiments
were performed on a VP MicroCalorimetry System. The experiment
was conducted by adding 2 (70 µM) every 3 min into a sodium
cacodylate buffer (10 mM, pH 7.0) containing NaCl (100 mM) and
11-mer G-G mismatch-containing duplex d(CTA ACG GAA TG)/
d(CAT TCG GTT AG) (GG2) (0 or 4 µM) at 25 °C.
N-(7-Methylpyridino[3,2-e]pyridin-2-yl)-3-((2-(N-(7-methylpyri-
dino[3,2-e]pyridin-2-yl)carbamoyl)ethyl)amino)propanoate (2). To
a solution of 8 (50 mg, 0.09 mmol) in ethyl acetate (3 mL) was added
ethyl acetate containing 4 M HCl (1.5 mL) at 0 °C, and the mixture
was stirred at room temperature for 2 h. The solvent was evaporated
to dryness to give the hydrochloride of 2 (quantitative yield) as a white
solid. The hydrochloride of 2 was dissolved in H2O and extracted into
CHCl3 by the addition of 28% aqueous ammonia solution. The organic
layer was dried over MgSO4, and the solvent was evaporated in vacuo
to give 2 (34 mg, 84%) as a pale yellow solid: 1H NMR (CDCl3, 400
MHz) δ 8.39 (d, 2 H, J ) 8.8 Hz), 8.03 (d, 2 H, J ) 8.8 Hz), 7.93 (d,
2 H, J ) 8.0 Hz), 7.21 (d, 2 H, J ) 8.0 Hz), 3.10 (t, 4 H, J ) 6.0 Hz),
2.77 (t, 4 H, J ) 6.0 Hz), 2.72 (s, 6 H); 13C NMR (CDCl3, 100 MHz)
δ 171.9, 163.0, 154.6, 153.7, 138.7, 136.3, 121.3, 118.4, 114.6, 44.6,
36.8, 25.5; FABMS (NBA), m/e 444 [(M + H)+]; HR-FABMS calcd
for C24H26O2N7 [(M + H)+], 444.2146; found, 444.2148.
4-((tert-Butoxy)carbonylamino)-N,N-bis(2-(N-(7-methylpyridino-
[3,2-e]pyridin-2-yl)carbamoyl)ethyl)butanamide (9). To a solution
of 2 (200 mg, 0.23 mmol) dissolved in DMF (5 mL) were added
pentafluorophenyl 4-(tert-butoxy)carbonylaminobutylate (100 mg, 0.27
mmol) and diisopropylethylamine, and the solution was stirred over-
night. Solvents were evaporated, and the residue was chromatographed
on a silica gel column (chloroform-methanol ) 50:1) to give 9 (223
1
mg, 87%); H NMR (400 MHz, CDCl3) δ 10.0 (broad, 1 Η), 9.54
(broad, 1 H), 8.37 (d, 2 H, J ) 8.9 Hz), 8.06 (d, 2 H, J ) 8.8 Hz), 7.94
(d, 1 H, J ) 8.1 Hz), 7.91 (d, 1 H, J ) 8.3 Hz), 7.22 (d, 1 H, J ) 8.2
Hz), 7.18 (d, 1 H, J ) 8.1 Hz), 5.21 (broad, 1 H), 3.76 (broad, 2 H),
3.66 (broad, 2 H), 3.08 (broad, 2 H), 2.82 (broad, 4 H), 2.69 (s, 6 H),
2.41 (t, 2 H, J ) 7.1), 1.78 (broad, 2 H), 1.35 (s, 9 H); 13C NMR (100
MHz, CDCl3) δ 173.2, 171.3, 170.2, 163.2, 163.0, 156.1, 153.5, 139.1,
139.0, 136.4, 136.3, 121.6, 121.4, 118.5, 118.4, 114.6, 78.9, 77.3, 77.2,
77.0, 76.7, 44.6, 42.1, 40.0, 36.8, 36.2, 30.2, 28.4, 25.2, 25.5, 25.3;
FABMS (NBA), m/e (%) 629 [(M + H)+] (40); HR-FABMS calcd for
C33H41O4N4F5, 629.3200; found, 629.3193.
4-Amino-N,N-bis(2-(N-(7-methylpyridino[3,2-e]pyridin-2-yl)car-
bamoyl)ethyl)butanamide (3). To a solution of 9 (220 mg, 0.35 mmol)
in chloroform (1 mL) was added 3 N HCl in AcOEt (5 mL) at 0 °C,
and the solution was stirred at room temperature for 30 min. The
resulting solution was evaporated. Water and aqueous ammonia were
added to the solution (pH > 7), and the resulting solution was extracted
with chloroform and dried with MgSO4. The solution was evaporated
to give 3 (167 mg, 90%). 1H NMR (400 MHz, CDCl3) δ 7.76 (broad,
1 Η), 7.68-7.61 (3 H), 7.54 (d, 1 H, J ) 9.0) 7.43 (d, 1 H, J ) 8.2),
7.00 (d, 1 H, J ) 8.2), 6.81 (d, 1 H, J ) 8.0), 3.63 (t, 2 H, J ) 5.9),
3.56 (t, 2 H, J ) 6.2), 2.95 (t, 2 H, J ) 7.5), 2.71-2.62 (7 H), 2.40 (s,
3 H), 2.35 (s, 3 H), 1.87 (p, 2 H, J ) 7.5); 13C NMR (100 MHz, CDCl3)
δ 183.1, 178.0, 176.0, 174.9, 166.0, 165.8, 155.1, 154.9, 142.4, 142.2,
140.1, 124.7, 124.5, 120.8, 116.6, 116.6, 47.4, 46.3, 41.8, 38.6, 38.4,
32.7, 26.6, 25.8, 25.2; FABMS (NBA), m/e (%) 529 [(M + H)+] (40);
HR-FABMS calcd C29H33N9O3, 529.2676; found, 529.2694.
Pentafluorophenyl-3-((tert-butoxy)-N-(2-((pentafulorophenyl)oxy-
carbonyl)ethyl)carbonylamino)propanoate (10). Compound 5 was
converted to N-(tert-butoxycarbonyl)iminobis(propionic acid) as de-
scribed above. To a solution of N-(tert-butoxycarbonyl)iminobis-
(propionic acid) (6.2 g, 24 mmol) in DMF were added pentafluorophe-
nol (10.5 g, 57 mmol) and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride (10.9 g, 57 mmol). The mixture was stirred for 24 h at
room temperature. The solvent was evaporated, and the crude product
was suspended in CHCl3. The organic layer was washed with H2O and
dried over MgSO4. The solvent was evaporated in vacuo, and the crude
residue was purified by silica gel column chromatography to give 9
(8.0 g, 57%) as a white solid: 1H NMR (CDCl3, 400 MHz) δ 3.65 (m,
4 H), 2.99 (m, 2 H), 2.93 (m, 2 H), 1.47 (s, 9 H); FABMS (NBA), m/e
594 [(M + H)+]; HR-FABMS calcd for C23H18O6NF10 [(M + H)+],
594.0973; found, 594.0971.
NMR Measurements of 2-GG1 Complex. DNA duplex GG1 (1
mM in H2O) both in free and complex states with 2 was pipetted into
a Shigemi microtube. The NMR experiments were performed on Bruker
DRX-600 and JEOL R-400 spectrometers. The total correlated spec-
troscopy (TOCSY) spectra were recorded with mixing time of 60 ms,
and nuclear Overhauser effect spectroscopy (NOESY) spectra were
recorded with mixing time of 300 ms. Total of 32 (TOCSY) and 64
(NOESY) transients were acquired with a recycle delay of 1.2 s at 17
°C. The 31P NMR spectra were taken with proton decoupling sequence
during acquisition time.
Acknowledgment. We thank Mr. Junji Nakamura and Mr.
Yuji Hirohata of Nihon SiberHegner Co., Ltd. for ITC measure-
ments. This research was supported in part by a grant-in-aid
for scientific research on priority areas (C) “medical genome
science” from the Ministry of Education, Science, Sports and
Culture of Japan.
Supporting Information Available: Thermal profile of
1
imino protons in H NMR of GG1 and the 2-GG1 complex
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
3-((tert-Butoxy)-N-(2-(N-(2-quinolyl)carbamoyl)ethyl)carbonyl-
amino)-N-(7-methylpyridino[3,2-e]pyridin-2-yl)propanamide (12).
JA0109186