Highly sensitive detection of GG mismatched DNA by surfaces immobilized naphthyridine dimer through poly(ethylene oxide) linkers

Article abstract of DOI:10.1016/j.bmcl.2003.12.079

Naphthyridine dimer is a unique molecule that strongly, and selectively, binds to the guanine-guanine mismatch in duplex DNA. We have synthesized naphthyridine dimers possessing a different length of poly(ethylene oxide) (PEO) linker, and immobilized them

Full text of DOI:10.1016/j.bmcl.2003.12.079

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1105–1108
Highly sensitive detection of GG mismatched DNA by
surfaces immobilized naphthyridine dimer through
poly(ethylene oxide) linkers
Kazuhiko Nakatani,a,b,* Akio Kobori,b Hiroyuki Kumasawaa and Isao Saitoa
aDepartment of Synthetic Chemistry and Biological Chemistry, Faculty of Engineering, Kyoto University, Kyoto 615-8510, Japan
bPRESTO, Japan Science and Technology Agency (JST), Kyoto 615-8510, Japan
Received 10 November 2003; revised 24 December 2003; accepted 25 December 2003
Abstract—Naphthyridine dimer is a unique molecule that strongly, and selectively, binds to the guanine–guanine mismatch in
duplex DNA. We have synthesized naphthyridine dimers possessing a different length of poly(ethylene oxide) (PEO) linker, and
immobilized them to CM5 sensor chip to carry out a surface plasmon resonance (SPR) assay of DNA duplexes containing a single
base mismatch. The sensitivity of the sensor remarkably increased with increasing numbers of PEO units incorporated into
the linker. With the sensor surface immobilized naphthyridine dimer for 1.5ꢀ103 response unit (RU) through three PEO units, the
distinct SPR signal was observed at a concentration of 1 nM of the 27-mer G–G mismatch.
# 2004 Elsevier Ltd. All rights reserved.
Discrimination of base mismatches from normal Wat-
son–Crick base pairs in duplex DNA constitutes a key
technology for the detection of single nucleotide poly-
morphisms (SNPs). To detect SNPs, a number of high
throughput methods have been developed.1ꢁ3 Most of
these technologies use the difference in thermodynamic
stability between mismatched and fully matched
duplexes. We,4ꢁ10 and others,11ꢁ13 have pursued an
alternative approach to SNPs detection by using small
molecular ligands that can selectively bind to the mis-
matched base pairs. This is a modified method of het-
eroduplex analyses, which determines the presence of
SNPs in testing sample DNAs by detecting the mis-
matched base pairs in heteroduplex produced between
sample and the standard DNAs.14,15 We have developed
a novel sensor for a surface plasmon resonance (SPR)
assay to detect the G–G mismatch duplexes by employ-
ing the surface where naphthyridine dimer 1 was
immobilized.4 Naphthyridine dimer 1 (Fig. 1) consisting
of two 2-amino-7-methyl-1,8-naphthyridines having
complementary surface of hydrogen bonding to a gua-
nine and a linker connecting two chromophores binds
selectively to the guanine–guanine (G–G) mismatch in
duplex DNA.4,5 In previous studies,4 the sensor surface
was synthesized by immobilizing 2 on a dextran matrix
pre-coated on the gold surface of a Biacore CM5 sensor
chip.16 The SPR signals obtained by the sensor surface
were much weaker than those expected from the mole-
cular mass of the analyte DNA. It was anticipated that
naphthyridine dimers immobilized with a short alkyl
linker (e.g., 2) would not be exposed on the surface, but
would be folded back into the matrix. We have inserted
hydrophilic poly(ethylene oxide) (PEO) units between 1
Keywords: Mismatch; SPR; PEO.
* Corresponding author. Tel.: +81-75-383-2756; fax: +81-75-383-
2759; e-mail: nakatani@sbchem.kyoto-u-ac.jp
Figure 1. Structures of naphthyridine dimers with a different
poly(ethylene oxide) linker length.
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.12.079

1106
K. Nakatani et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1105–1108
and the dextran matrix with the expectation that it
would suppress undesirable folding, and so increase the
fraction of ligand exposed on the surface. We here
report that a new sensor surface of naphthyridine dimer
having a long PEO spacer in the linker remarkably
increased the sensitivity to the G–G mismatch 13times
that of the prototype. With the sensor surface immobi-
lized naphthyridine dimer for 1.5ꢀ103 response unit
(RU) through three PEO units, the distinct SPR signal
was observed at a concentration of 1 nM of the 27-mer
G–G mismatch.
terminal attached to the dextran surface according to
the procedure recommended by Biacore (Scheme 2).16,19
We synthesized two 5-immobilized sensor surfaces
(Sensors 1 and 2), which differed in their ligand densities
on the surface, one 4-immobilized sensor surface (Sen-
sor 3), and one 3-immobilized sensor surface (Sensor 4).
SPR detects changes in the refractive index on the sur-
face layer caused by variation of the mass on the sensor
chip surface, for example, when the analyte binds to the
immobilized ligand on the surface. The change in SPR
signal, termed the SPR response presented in response
units (RU), is directly related to the change in surface
concentration of biomolecules. SPR response of 1000
RU is equivalent to the change in surface concentration
of 1 ng/mm2. Thus, the density of immobilized ligands
and analyte bound on the surface could be calculated by
the difference in SPR response before and after the
Naphthyridine dimers tethered to a PEO linker were
synthesized as shown in Scheme 1. First, an activated
amino acid of a PEO unit 10 was synthesized from tet-
raethylene glycol. Though the PEO unit related to 10
had been synthesized from triethylene glycol,17,18 it was
conveniently obtained by oxidation of mono-protected
tetraethylene glycol. Mono mesylation of tetra(ethylene
glycol) and a subsequent reaction with sodium azide
produced o-azide alcohol 6, which was oxidized with
Jones’ reagent to carboxylic acid 7. Hydrogenation of
the azide 7 followed by protection of the resulting pri-
mary amino group in 8 produced N-Boc-amino acid 9.
The carboxylic acid of 9 was activated as a form of
pentafluoro phenyl ester, giving the PEO unit 10.
Naphthyridine dimer tethered to an amino alkyl linker
24 was reacted with 10. The N-Boc group on the amino
termini of the resulting 11 was deprotected to afford the
naphthyridine dimer 3 having one PEO unit. The pri-
mary amine 3 was coupled with 10 giving 12, which was
deprotected to form 4 having two PEO units. Repetition
of the coupling–deprotection sequence provided 5 hav-
ing three PEO units.
analysis.16 Sensor 1 was prepared by immobilizing 5 for
2
5.5ꢀ10 RU (1 RU=1 pg mmꢁ2, and it had a ligaꢁn2d
.
density on the sensor surface of 4.7ꢀ10ꢁ10 fmol nm
.
.
Sensor 2 was prepared by immobilizing 5 for 1.5ꢀ103
RU. The density of 5 on the surface was about
1.3ꢀ10ꢁ9 fmol nmꢁ2. Sensors 3 and 4 were prepared by
.
immobilizing 4 and 3 for 4.6ꢀ102 and 5.4ꢀ102 RU,
respectively. The ligand density of Sensors 3 and 4 were
, respectively,
which are similar to the ligand density of Sensor 1.
4.8ꢀ10ꢁ10 and 7.0ꢀ10ꢁ10 fmol nm
ꢁ2
.
The DNA oligomers used for the SPR studies were 27-
mer duplex 50-d(GTT ACA GAA TCT XYZ AAG
CCT AAT ACG)-30/30-d(CAA TGT CTT AGA X0Y0Z0
TTC GGA TTA TGC)-50 (1 mM) containing a G–G
mismatch (XYZ/X0Y0Z0=CGG/GGC represented a
DNA duplex having the trinucleotide block) in the
middle of the sequence. The effect of the linker length
on the SPR intensity was assessed by measuring the
specific binding of CGG/GGC to Sensors 1, 3, and 4, on
Naphthyridine dimers tethered to a different length of
PEO linker were immobilized on an activated carboxyl
Scheme 1. Synthesis of naphthyridine dimers tethered to a different
PEO linker. Reagents and conditions: (a) MsCl, NEt3, dry diethyl-
ether; (b) NaN3, EtOH, reflux, for 2 steps 46%; (c) Jones’ reagent,
acetone, 80%; (d) Pd–C, H2, EtOH, quantitative; (e) (Boc)2O, 2 M,
NaOH, THF, 36%; (f) pentafluorophenol, EDCI, DMF, 80%; (g) 10,
Et3N, THF, 90% for 11, 78% for 12, 74% for 13; (h) HCl, AcOEt,
CHCl3, quantitative.
Scheme 2. Immobilization of naphthyridine dimers having different
ligand density (LD, fmol nmꢁ2) and lengths of a PEO linker on to the
.
dextran coated gold surface.

K. Nakatani et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1105–1108
1107
which the naphthyridine dimer was covalently bound
through 3, 2, and 1 PEO units, respectively. The buffer
used for the analysis was 10 mM of HEPES (pH=7.4)
containing 150 mM of NaCl. The measurements were
carried out at 25 ꢂC. Because the degree of immobiliza-
tion of the ligand on the surface was different for Sen-
sors 1, 3, and 4, the SPR intensity was normalized, and
was quoted in units of mass of DNA bound to a unit
per immobilized ligands (pg of DNA per pmol of an
immobilized ligand) (Fig. 2). A prototype of the sensor
where 2 ꢁw1as immobilized showed a signal intensity of 94
concentration of 1 nM for the 27-mer duplex was found
to be the lower limit for the detection of the G–G mis-
match. Because the SPR intensity increased with
increasing molecular weight of the analyte, the detection
limit of longer duplexes containing the G–G mismatch
would be below 1 nM.
The effect of the sequence flanking to the G–G mis-
match on the SPR detection by the 1-immobilized sur-
face was investigated by analyzing 10 G–G mismatches
by Sensor 1. The sensorgrams were obtained by apply-
ing 1 mM of XYZ/X0Y0Z0 to Sensor 1. The SPR signals
for all the G–G mismatches were clearly detected and
.
pg pmol after a period of 180 s after the injection of
CGG/GCC. As can be seen from Figure 2, the SPR
intensityꢁ1increased to 1.5ꢀ102, 4.9ꢀ102, and 1.2ꢀ103
.
pg pmol
as increasing the number of PEO units
incorporated in the linker from one to three. The signal
obtained for Sensor 1 is about 13-fold stronger in
intensity than that obtained for 2-immobilized sensor.
Further incorporation of the PEO unit in the linker did
not produce remarkable signal enhancement.
Having found that the incorporation of three PEO units
significantly increased the SPR intensity, we then deter-
mined the detection limit of the G–G mismatch duplex
by using Sensor 2. A distinct SPR signal was observed
for CGG/GGC at a concentration of 1 nM, whereas
duplexes containing a G–A mismatch (XYZ/X0Y0Z0=
CGG/GAC), a G–T mismatch (XYZ/X0Y0Z0=CGG/
GTC), and a G–C match (XYZ/X0Y0Z0=CGG/GCC)
base pair did not show any significant signal (Fig. 3).
Further lowering of the concentration of CGG/GGC
resulted in a loss of signal. By using a refined linker with
three PEO units for the immobilization of the naph-
thyridine dimer and the Biacore 2000 instrument, a
Figure 3. Detection of mismatched DNAs at a concentration of 1 nM
by Sensor 2. Key: (a) CGG/GGC, (b) CGG/GAC, (c) CGG/GTC, and
(d) CGG/GCC.
Figure 2. Binding of 27-mer duplex 50-d(GTT ACA GAA TCT XYZ
AAG CCT AAT ACG)-30/30-d(CAA TGT CTT AGA X0Y0Z0 TTC
GGA TTA TGC)-50 (1.0 mM) containing the G–G mismatch (XYZ/
X0Y0Z0=CGG/GGC) to the naphthyridine dimer-immobilized sensors
(Sensors 1, 3, and 4) possessing a different length of PEO unit. Key: (a)
Sensor 1, (b) Sensor 3, (c) Sensor 4, and (d) 2-immobilized sensor. The
signal intensity represents an amount of DNA (pg) bound to a unit
amount (pmol) of each immobilized ligand on the surface.
Figure 4. Sensorgrams obtained for the binding of XYZ/X0Y0Z0 to
Sensor 1 at a concentration of 1 mM duplex. Binding was measured for
180 s and dissociation for 120 s. Key: (a) GGC/CGG, (b) CGG/GGC,
(c) CGC/GGG, (d) TGG/AGC, (e) TGA/AGT, (f) AGG/TGC, (g)
TGC/AGG, (h) AGC/TGG, (i) AGA/TGT, (j) AGT/TGA, (k) CGG/
GAC (GA mismatch).

1108
K. Nakatani et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1105–1108
sequence dependent on the flanking base pairs (Fig. 4).
The strongest SPR signal was observed for the GGC/
CGG, whereas the weakest signal was observed for
AGT/TGA. The difference in SPR intensity between
these two G–G mismatches is about 6 folds. The G–G
mismatches flanking G–C base pairs binds to the 1-
immobilized surface stronger than those flanking A–T
base pairs. These observations are in good agreement
with the results obtained by DNase I foot print titration
and UV-melting studies of duplexes containing the G–G
mismatch in the presence of 1.15 In contrast, the SPR
signal observed for the G–A mismatch was quite weak
as compared to those of the G–G mismatch. These
results showed that Sensor 1 is capable of detecting the
G–G mismatch in duplex DNA irrespective of the
sequence flanking to the mismatch.
References and notes
1. Syvanen, A.-C. Nat. Rev. Gen. 2001, 2, 930.
¨
2. Kwok, P. Y. Annu. Rev. Genom. Hum. G 2001, 2, 235.
3. Schafer, A. J.; Hawkins, J. R. Nat. Biotechnol. 1998, 16, 3 3 .
4. Nakatani, K.; Sando, S.; Saito, I. Nat. Biotechnol. 2001,
19, 51.
5. Nakatani, K.; Sando, S.; Kumasawa, H.; Kikuchi, J.;
Saito, I. J. Am. Chem. Soc. 2001, 123, 12650.
6. Nakatani, K.; Sando, S.; Saito, I. Bioorg. Med. Chem.
2001, 9, 2381.
7. Smith, E. A.; Kyo, M.; Kumasawa, H.; Nakatani, K.;
Saito, I.; Corn, R. M. J. Am. Chem. Soc. 2002, 124, 6810.
8. Nakatani, K.; Hagihara, S.; Sando, S.; Sakamoto, S.;
Yamaguchi, K.; Maesawa, C.; Saito, I. J. Am. Chem. Soc.
2003, 125, 662.
9. Kobori, A.; Horie, S.; Suda, H.; Saito, I.; Nakatani, K. J.
Am. Chem. Soc. 2004, 126, 557.
10. Hagihara, S.; Kumasawa, H.; Goto, Y.; Hayashi, G.;
Kobori, A.; Saito, I.; Nakatani, K. Nucleic Acids Res.
2004, 32, 278.
11. Jackson, B. A.; Barton, J. K. J. Am. Chem. Soc. 1997,
119, 12986.
12. Jackson, B. A.; Alekseyev, V. Y.; Barton, J. K. Biochem-
istry 1999, 38, 4655.
13. Lacy, E. R.; Cox, K. K.; Wilson, W. D.; Lee, M. Nucleic
Acids Res. 2002, 30, 1834.
14. Landers, J. P. Anal. Chem. 2003, 75, 2919.
15. Nataraj, A. J.; Olivos-Glander, I.; Kusukawa, N. Elec-
trophoresis 1999, 20, 1177.
16. Pharmacia-Biosensor. 1990 Biacore User0s Manual. Pis-
catway, NJ.
Unlike most SPR studies using the surfaces where mac-
romolecules were immobilized, the drug-immobilized
surface was extremely sensitive to the way of drug
immobilization due to a small molecular size relative to
the surface and its hydrophobic character. The studies
described here reveal that incorporation of three PEO
linker units between naphthyridine dimer and the dex-
tran surface is especially effective in increasing the SPR
sensitivity without losing the mismatch specificity.
Acknowledgements
17. Jeong, S. W.; O’Brien, D. F. J. Org. Chem. 2001, 66, 4799.
18. Boumrah, D.; Campbell, M. M.; Fenner, S.; Kinsman,
R. G. Tetrahedron 1997, 53, 6977.
This work was supported by Grant-in-Aid for Scientific
Research on Priority Areas (C) ‘Medical Genome
Science’ from the Ministry of Education, Culture,
Sports, Science and Technology of Japan.
19. Lofa
˚
s, S.; Johnson, B. J. Chem. Soc., Chem. Commun.
¨
1990, 21, 1526.