Cyclohexyl "base pairs" stabilize duplexes and intensify pyrene fluorescence by shielding it from natural base pairs

Article abstract of DOI:10.1039/c1ob06325a

In this study, we investigated the stability and structure of artificial base pairs that contain cyclohexyl rings. The introduction of a single pair of isopropylcyclohexanes into the middle of DNA slightly destabilized the duplex. Interestingly, as the number of the "base pairs" increased, the duplex was remarkably stabilized. A duplex with six base pairs was even more stable than one containing six A-T pairs. Thermodynamic analysis revealed that changes in entropy and not enthalpy contributed to duplex stability, demonstrating that hydrophobic interactions between isopropyl groups facilitated the base pairing, and thus stabilized the duplex. NOESY of a duplex containing an isopropylcyclohexane-methylcyclohexane pair unambiguously demonstrated its "pairing" in the duplex because distinct NOEs between the protons of cyclohexyl moieties and imino protons of both of the neighboring natural base pairs were observed. CD spectra of duplexes tethering cyclohexyl moieties also showed a positive-negative couplet that is characteristic of the B-form DNA duplex. Taken together, these results showed that cyclohexyl moieties formed base pairs in the DNA duplex without severely disturbing the helical structure of natural DNA. Next, we introduced cyclohexyl base pairs between pyrene and nucleobases as an "insulator" that suppresses electron transfer between them. We found a massive increase in the quantum yield of pyrene due to the efficient shielding of pyrene from nucleobases. The cyclohexyl base pairs reported here have the potential to prepare highly fluorescent labeling agents by multiplying fluorophores and insulators alternately into DNA duplexes. The Royal Society of Chemistry 2011.

Full text of DOI:10.1039/c1ob06325a

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 8313
www.rsc.org/obc
PAPER
Cyclohexyl “base pairs” stabilize duplexes and intensify pyrene fluorescence
by shielding it from natural base pairs†
Hiromu Kashida,* Koji Sekiguchi, Naofumi Higashiyama, Tomohiro Kato and Hiroyuki Asanuma*
Received 4th August 2011, Accepted 5th October 2011
DOI: 10.1039/c1ob06325a
In this study, we investigated the stability and structure of artificial base pairs that contain cyclohexyl
rings. The introduction of a single pair of isopropylcyclohexanes into the middle of DNA slightly
destabilized the duplex. Interestingly, as the number of the “base pairs” increased, the duplex was
remarkably stabilized. A duplex with six base pairs was even more stable than one containing six A–T
pairs. Thermodynamic analysis revealed that changes in entropy and not enthalpy contributed to
duplex stability, demonstrating that hydrophobic interactions between isopropyl groups facilitated
the base pairing, and thus stabilized the duplex. NOESY of a duplex containing an
isopropylcyclohexane–methylcyclohexane pair unambiguously demonstrated its “pairing” in the duplex
because distinct NOEs between the protons of cyclohexyl moieties and imino protons of both of the
neighboring natural base pairs were observed. CD spectra of duplexes tethering cyclohexyl moieties
also showed a positive–negative couplet that is characteristic of the B-form DNA duplex. Taken
together, these results showed that cyclohexyl moieties formed base pairs in the DNA duplex without
severely disturbing the helical structure of natural DNA. Next, we introduced cyclohexyl base pairs
between pyrene and nucleobases as an “insulator” that suppresses electron transfer between them. We
found a massive increase in the quantum yield of pyrene due to the efficient shielding of pyrene from
nucleobases. The cyclohexyl base pairs reported here have the potential to prepare highly fluorescent
labeling agents by multiplying fluorophores and insulators alternately into DNA duplexes.
Recently, Leumann et al. incorporated a non-planar artificial base,
1. Introduction
cyclohexylbenzene, into the middle of oligodeoxyribonucleotides
(ODN) via D-ribose.23 Surprisingly, these base pairs greatly
Natural DNA forms a stable double-helical structure, both by
hydrogen bonding and aromatic stacking interactions. Natural
stabilized a DNA duplex through a cyclohexyl/phenyl interaction,
base pairs achieve high stability and orthogonality by controlling
suggesting that even non-planar molecules can be accommodated
the strength of these two interactions in a very sophisticated
manner. In 1995, Kool et al. first reported that hydrogen bonding is
not a prerequisite for the stability of the DNA duplex; an artificial
base pair, which has aromatic rings but no hydrogen bonding
sites, can stabilize the DNA duplex.1 Since then, a wide variety of
artificial base pairs that use aromatic stacking interactions and/or
hydrogen bonding has been reported.2–9 In addition, non-ribose
scaffolds were also utilized to design various base-surrogates
involving functional molecules.10–17 These artificial nucleotides
have the potential to serve as novel nanomaterials and biological
tools.18–22 Furthermore, investigation into the tolerance of duplex
stability by changing the structure of bases and linkers should
provide insight as to why nature selected DNA as genetic carriers.
inside the DNA duplex. However, to our knowledge, artificial
base pairs without aromatic rings have not been reported on so
far, because aromatic rings have been believed to be essential to
stabilize the duplex by stacking or CH–p interactions.
Previously, we reported novel “base pairs” tethering cyclohexyl
moieties (H and I residues in Fig. 1b) on D-threoninol as an
insulator that inhibits electron or hole transfer.24 When the base
pairs were introduced between nucleobases and perylenediimide
(PDI), the quantum yield of PDI greatly increased. Concurrently,
we found that base pairs of isopropylcyclohexane moieties (H) ad-
jacent to PDI moieties stabilize the whole DNA duplex as the num-
ber of base pairs increase, even though they have neither hydrogen
bonds nor aromatic stacking interactions. These results prompted
us to investigate the base pairs of cyclohexyl base surrogates in
detail. If non-aromatic molecules do not destabilize DNA and thus
work as artificial bases, the molecules available to be incorporated
into DNA should be extended, allowing for further functional-
ization of DNA. In this study, we first analyzed the structure
and the stability of duplexes containing a cyclohexyl base pair in
detail. Next, we applied the base pair as an insulator to enhance
Graduate School of Engineering, Nagoya University, Furocho,
Chikusa, Nagoya, 464-8603, Japan. E-mail: kashida@mol.nagoya-
u.ac.jp, asanuma@mol.nagoya-u.ac.jp; Fax: +81-52-789-2528; Tel: +81-
52-789-2488
† Electronic supplementary information (ESI) available: Synthetic proce-
dures of methylcyclohexane moeity, 2D NOESY between aliphatic protons
and UV-Vis spectra. See DOI: 10.1039/c1ob06325a
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8313–8320 | 8313
©

View Article Online
Fig. 1 (a) A schematic representation of duplexes modified with cyclohexyl base pair(s). (b) Sequences of the modified ODNs synthesized in this study.
The chemical structures of insulators and fluorophore moieties are also shown.
pyrene emission, since this artificial base pair can suppress electron
transfer to and from nucleobases due to the lack of p electrons.
correspond to six natural base pairs, were observed in the imino
proton region, although the signals of imino protons of G4 and G7
overlapped. These results indicated that incorporation of a non-
planar base pair did not severely disturb the base pairing of natural
nucleobases. NOEs between imino protons and protons of H and
K are also shown in Fig. 2. Methyl protons of H showed NOEs
to imino protons of both of G4 and T10. Similarly, distinct NOEs
were observed between methyl protons of K and imino protons of
G4 and T10. These NOEs unambiguously demonstrated that both
the isopropyl group of H and the methyl group of K were located
adjacent to both G4 and T10. In addition, Hb of H showed NOE
with an imino proton of T10, whereas Ha of K did so with an imino
proton of G4, suggesting that the orientations of the cyclohexane
rings of H and K are reversed. Intermolecular NOEs between H
and K were also observed, showing that H and K are located in
close proximity (see Fig. S1 in Supporting Information†). NOESY
definitely demonstrated that H and K formed a “base pair” inside
the DNA duplex. Computer modeling without any restraints
by Insight II/Discover 3 was also consistent with the results of
NOESY (Fig. 3). Thus, we concluded that cyclohexyl moieties
worked as an artificial base pair; the moieties were located inside
2. Results
2.1 Structural analysis of a duplex containing artificial base pairs
We initially investigated the structure of a duplex containing a
cyclohexyl base pair. If non-natural molecules work as a base pair,
molecules should be located inside the DNA duplex and interact
with each other. To confirm its solution structure, NMR structural
analysis was conducted with a 7mer duplex (H1C/K1D in Fig. 1)
tethering an H–K pair at its center position. A K residue was
introduced into DNA via D-threoninol25 using phosphoramidite
chemistry (see Supporting Information†). In order to monitor
imino protons that are exchangeable with water molecules, NMR
was measured in H2O (H2O/D2O, 9 : 1) at 275 K with a 3-9-19
WATERGATE pulse sequence for H2O suppression.26 Most of the
signals of the duplex, except for the protons of the cyclohexane
rings, could be assigned from the NOESY, DQFCOSY and
TOCSY spectra. The one-dimensional NMR chart of the imino
proton region is shown at the top of Fig. 2. Six peaks, which
Fig. 2 2D NOESY spectra of H1C/K1D between the aliphatic protons
and imino protons in H2O/D2O (9/1) at 275 K. 1D spectra of the imino
protons and aliphatic protons are shown at the top and left of the chart,
respectively. The residue numbers and proton numbers of H and K residues
are shown at the left.
Fig. 3 Energy-minimized structure of H1C/K1D that contains an H–K
pair calculated by Insight II/Discover 3. The H and K residues are colored
in yellow and orange, respectively.
8314 | Org. Biomol. Chem., 2011, 9, 8313–8320
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 1 Thermodynamic parameters of duplexes containing artificial
the duplex and interacted with each other. It should be noted that
the isopropyl group of H and the methyl group of K are in close
proximity, indicating that these groups interact with each other.
Next, we measured the CD spectra of duplexes containing
H–H pairs in order to investigate the structure of a DNA
duplex tethering multiple cyclohexyl base pairs, as shown in
Fig. 4. H2A/H2B carrying two H–H pairs showed a similar
positive couplet to that of a native duplex (P0/N). Similarly,
the incorporation of six H–H pairs did not significantly alter
the spectra; H6A/H6B showed similar spectra to those of other
duplexes. These results indicated that the incorporation of H–H
pairs did not disturb the basic B-form structure of nucleobases.
From NMR and CD analyses, we concluded that cyclohexyl
moieties worked as base pairs without disturbing the B-form
structure, although these bases had non-planar structures.
base pairs
Tm/ -DH/
-DS/
-DG◦
/
37
Sequence
◦Ca kcal mol-1 cal-1 K-1 mol-1 kcal mol-1
H1A/H1B
44.7
86.2
245
10.1
I1A/I1B
49.5
47.0
41.1
47.1
93.3
83.6
79.0
86.9
265
236
226
246
11.2
10.5
8.8
J1A/J1B
K1A/K1B
H2A/H2B
10.7
K2A/K2B
H6A/H6B
40.3
61.3 102.0
81.8
235
279
8.8
15.4
K6A/K6B
P0/Nb
47.4
47.7
49.4
53.4 100.5
49.2 101.6
53.6 134.2
85.4
89.9
94.0
241
254
266
282
289
385
10.7
11.2
11.6
13.1
11.9
14.9
A1A/T1Bb
G1A/C1Bb
A2A/T2B
A6A/T6B
a Solution conditions: [ODN] = 5.0 mM, [NaCl] = 100 mM, pH 7.0 (10 mM
phosphate buffer) b Data from ref. 27.
effect of H–H with a conventional natural base pair.27 A1A/T1B
and G1A/C1B gave higher Tms than H1A/H1B. Similarly, the Tm
of A2A/T2B was higher than that of H2A/H2B. However, the
Tm of six A–T pairs (A6A/T6B: 53.6 ◦C) was lower than that of
six H–H pairs (H6A/H6B: 61.3 ◦C); H–H pairs were more stable
than A–T pairs when multiple H–H pairs were introduced. The
Tms of other artificial base pairs are also listed in Table 1. On the
other hand, artificial bases carrying aromatic moieties stabilized
the duplex in spite of the incorporation of a single pair; Tms of
I1A/I1B and J1A/J1B containing a benzene ring were 49.5 and
47.0 ◦C, respectively, which were comparable or even higher than
that of the native duplex.
Fig. 4 CD spectra of P0/N, H2A/H2B, and H6A/H6B at 20 ◦C.
Conditions: [NaCl] = 100 mM, pH 7.0 (10 mM phosphate buffer), [ODN] =
5.0 mM.
2.2 Stability of duplexes containing artificial base pairs
We next investigated the thermal stability of duplexes containing
cyclohexyl base pairs. Here, we selected four kinds of molecules
as artificial bases via D-threoninol (H, I, J and K in Fig. 1b):
H–H and K–K homo-pairs do not carry benzene rings, but have
only cyclohexane rings. The I–I pair contains both benzene and
cyclohexane rings, whereas J–J contains no cyclohexane rings but
two benzene rings. By comparing the melting temperatures (Tms)
of duplexes involving these base pairs, the effect of aromatic rings
on duplex stability was evaluated.
The thermodynamic parameters of these duplexes determined
from 1/Tm versus ln(CT/4) plots are listed in Table 1. The -DG◦
37
value of H1A/H1B was 10.1 kcal mol-1, which was less than
that of P0/N (11.2 kcal mol-1). However, when the number of
H–H pairs increased to two, -DG◦ increased by 0.6 kcal mol-1
37
(H2A/H2B). Furthermore, -DG◦ of H6A/H6B was as high as
37
15.4 kcal mol-1, which was 4.2 kcal mol-1 higher than that of P0/N
without artificial base pairs. In addition, the -DG◦ was even 0.5
37
kcal mol-1 larger than that of A6A/T6B. The larger -DG◦ of
37
When a single H–H pair was introduced into DNA, its Tm
slightly dropped compared with the native duplex; the Tm of
H1A/H1B was 44.7 ◦C, whereas that of the native duplex (P0/N)
H6A/H6B was mainly attributed to the -DS, which was 106 cal
mol-1 K smaller than that of A6A/T6B, whereas its -DH was 32.2
kcal mol-1 smaller. These results strongly suggested that a smaller
loss in entropy due to hydrophobic interaction between H moieties
contributed to the large Tm.
◦
was 47.7 C (Table 1). The K–K pair that also lacks an aromatic
ring (K1A/K1B) further lowered its Tm to 41.1 ◦C, indicating
that cyclohexyl moieties destabilized DNA due to its non-planar
structure. However, as the number of the H–H pairs increased, Tm
did not decrease but somewhat increased. The Tm of H2A/H2B,
which contains two H–H pairs, was 2.4 ◦C higher than that
of H1A/H1B. Surprisingly, the incorporation of six H–H pairs
massively stabilized the DNA duplex; the Tm of H6A/H6B was as
high as 61.3 ◦C, which was 13.6 ◦C higher than that of P0/N. We
also measured Tms of native duplexes that contained A–T or G–C
pairs instead of H–H pairs in order to compare the stabilizing
The -DG◦ of the base pairs synthesized here was in the order
37
of I–I (11.2 kcal mol-1 ª P0/N) > J–J (10.5) ª H–H (10.1) >
K–K (8.8), which did not necessarily coincide with the number
of benzene rings in the base-surrogate. For example, the H–H
pair with isopropylcyclohexane was almost comparable with the
J–J pair having a biphenyl. Stability of a K–K pair was 1.3 kcal
mol-1 lower than that of an H–H pair. Furthermore, the difference
of -DG◦ between H6A/H6B and K6A/K6B was as high as 4.7
37
kcal mol-1, demonstrating that hydrophobic interactions between
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8313–8320 | 8315
©

View Article Online
Table 2 Spectroscopic behaviors and thermodynamic stabilities of the
modified ODNs
isopropyl groups contributed strongly to the stabilization of the
duplex.
Relative
lmax
/
Tm/
◦Cd
2.3 Shielding efficiency of cyclohexyl base pairs as insulators
Sequence
Intensitya
Ub
nmc
Next, we applied these cyclohexyl base pairs to suppress electron
transfer in DNA. Natural DNA nucleobases are known to be
good mediators for hole/electron transfer.28–32 However, since the
cyclohexyl base pair reported here has no p electrons, it should be
a good “insulator” of electron transfer when it is located next to
natural nucleobases. In order to evaluate the insulating ability of
cyclohexyl base pairs, we investigated the fluorescence recovery of
pyrene by the incorporation of cyclohexyl moieties. DNA–pyrene
conjugates have been widely used as fluorescence probes and
labeling agents.14–16,33–38 We have also introduced pyrene between
natural base pairs and prepared fluorescence probes and labeling
agents.39–41 However, quenching by nucleobases severely limits the
detection sensitivity; the quantum yield of pyrene is as low as
0.003, especially when multiple pyrenes are introduced between
natural base pairs.41 Since this quenching is attributable to electron
transfer from pyrene to nucleobases,33 we incorporated cyclohexyl
base pairs between pyrene and nucleobases as insulators to
enhance the emission of pyrene, as we previously applied to
perylenediimide24 (see Fig. 1). Kool et al. reported that 5,6-
dihydro-2¢-deoxythymidine, which they inserted between pyrene
and thymidine, enhanced the emission of pyrene in the single-
stranded state.42 In our study, we utilized the base pairing of
cyclohexyl moieties in order to suppress the undesired dynamic
quenching. Fluorescence enhancement by cyclohexyl base pairs
has the potential to improve the sensitivity of pyrene-based
fluorescent probes and labeling agents.
H2AP/H2B
1
0.19
346
48.5
e
I2AP/I2B
J2AP/J2B
K2AP/K2B
P1/N
0.70
<0.01
0.91
—
—
—
347
347
346
351
52.2
52.6
43.1
48.6
e
e
<0.01
<0.01
e
P1
0.02
—
350
—
e
H2AP
0.16
0.07
—
348
348
—
H2AP/N
0.02
43.2
P1/H2B
0.22
1.71
0.05
0.32
348
345
47.5
62.0
H6AP/H6B
e
I6AP/I6B
K6AP/K6B
H6AP
1.49
1.65
0.67
—
—
—
346
345
346
63.8
49.8
—
e
e
a Emission intensity at 378 nm relative to that of H2AP/H2B. Solution
conditions: [ODN] = 1.0 mM, [NaCl] = 100 mM, pH 7.0 (10 mM phosphate
buffer) b Quantum yield determined from the quantum yield of pyrene in
N2-bubbled cyclohexane (0.65) used as a reference. c Absorption maximum
of pyrene in the UV-Vis spectrum at 20 ◦C. d Solution conditions: [ODN] =
5.0 mM, [NaCl] = 100 mM, pH 7.0 (10 mM phosphate buffer) e Not
determined.
Sequences of modified DNA are shown in Fig. 1b and
schematically illustrated in Fig. 1a. We incorporated two or
six cyclohexyl base pairs (H–H and K–K pairs) between the
pyrene and nucleobases. In addition, I–I and J–J pairs were also
introduced in order to evaluate the effect of chemical structure on
the insulating ability. P1/N, which contains no cyclohexyl base
pairs, but a pyrene moiety, was synthesized as a control. P1/N
showed almost no fluorescence, as shown in the purple line in Fig.
5, because the emission from pyrene was severely quenched by
nucleobases. The quantum yield of pyrene was determined to be
below 0.01 (Table 2).43 On the other hand, incorporation of H–H
pairs greatly increased its emission. H2AP/H2B, which contains
two H–H pairs between pyrene and the neighboring base pairs,
showed intense peaks at around 380 and 400 nm (red line in Fig.
5). The quantum yield of H2AP/H2B increased to 0.19 (Table 2).
Consequently, the incorporation of two H–H pairs increased the
quantum yield of pyrene by more than hundred times. In contrast,
such a large enhancement of the quantum yield was not observed
with single-stranded H2AP; the emission intensity of H2AP was
about six times lower than that of H2AP/H2B (see Fig. 5 and
Table 2). In addition, H2AP/N and P1/H2B, which contained
two H moieties in one strand, showed much weaker emission than
H2AP/H2B. Thus, the base pairing of H moieties is required for
the large enhancement of the fluorescence intensity of pyrene.
We next evaluated the effect of the chemical structure on the
shielding effect. K–K pairs showed an almost comparable shielding
effect to that of H–H pairs (compare K2AP/K2B with H2AP/H2B
in Fig. 6 and Table 2). Accordingly, the isopropyl group is not es-
sential for the shielding effect, although it remarkably contributed
to the stability of the duplex. Although I2AP/I2B, containing
cyclohexylbenzene, exhibited a slightly lower emission intensity
than H2AP/H2B (compare the blue line with the red in Fig. 6),
the emission intensity was still remarkably higher than P1/N,
showing that introduction of one benzene ring did not severely
decrease the emission intensity. However, incorporation of the
Fig.
5
Fluorescence emission spectra of H2AP/H2B, H2AP/N,
P1/H2B, single-stranded H2AP and P1/N at 20 ◦C. Conditions: [NaCl] =
100 mM, pH 7.0 (10 mM phosphate buffer), [ODN] = 1.0 mM.
8316 | Org. Biomol. Chem., 2011, 9, 8313–8320
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig.
6
Fluorescence emission spectra of H2AP/H2B, I2AP/I2B,
Fig. 7 UV-Vis spectra of H2AP/H2B, H6AP/H6B and P1/N at 20 ◦C.
Conditions: [NaCl] = 100 mM, pH 7.0 (10 mM phosphate buffer), [ODN] =
5.0 mM.
J2AP/J2B, K2AP/K2B, H6AP/H6B, I6AP/I6B, and K6AP/K6B at
20 ◦C. Conditions: [NaCl] = 100 mM, pH 7.0 (10 mM phosphate buffer),
[ODN] = 1.0 mM.
DNA duplex, as revealed by NMR analysis in the case of
H–K “base pair”. Since a non-planar H moiety and planar
natural nucleobase should cause steric hindrance, hydrophobic
interactions between isopropyl groups should not compensate
for the destabilization caused by the disturbance of aromatic
stacking interactions between nucleobases. Consistent with this,
I1A/I1B or J1A/J1B, which involve aromatic rings, have higher
Tms due to aromatic stacking interactions, as well as hydrophobic
interactions.23 However, the accumulation of non-planar H–H
pairs did not further destabilize the duplex, but largely stabilized
J–J pairs with biphenyl moieties substantially lowered the emis-
sion intensity; J2AP/J2B emitted almost the same fluorescence
as P1/N did. From these comparisons, we concluded that at least
one cyclohexane ring is necessary in order to enhance the emission
of pyrene.
Further increases in the number of cyclohexyl base pairs mod-
erately enhanced the fluorescence. H6AP/H6B, which contained
six H–H pairs between the pyrene and nucleobases, showed a
1.7-times higher emission than H2AP/H2B (compare the red line
with the purple in Fig. 6) and its quantum yield was as high as
0.32 (Table 2). On the other hand, the emission intensity of single-
stranded H6AP was only about one-third of that of H6AP/H6B.
Similarly, incorporation of six I–I pairs or K–K pairs also enhanced
the fluorescence intensity, although the intensity of I6AP/I6B was
slightly lower than those of H6AP/H6B and K6AP/K6B (Fig. 6
and Table 2).
the duplex. In addition, -DG◦ did not proportionally increase
37
with the number of H–H pairs; the increase in -DG◦37 per H–H pair
between H2A/H2B and H6A/H6B was 1.17 kcal mol-1, while it
was only 0.6 kcal mol-1 between H1A/H1B and H2A/H2B. Such a
steep stabilization of H6A/H6B is attributable to the accumulated
hydrophobic interactions among the consecutive H moieties with
less steric hindrance between them. All of the cyclohexyl moieties
with chair-shaped configurations may be well accommodated
in the duplex and overcompensate the destabilization induced
at the boundary of H–H and natural base pairs. In addition,
hydrophobic interactions between isopropyl groups in the H–H
pair are also significant because K–K pairs with a methyl group on
the cyclohexyl ring less efficiently stabilized the duplex. Taken
together, we conclude that hydrophobic interactions without
aromatic stacking interactions can be a strong driving force of
DNA hybridization.
The melting temperatures of these duplexes are summarized in
Table 2. The Tms of duplexes sandwiched with two insulator pairs
were in the following order: J2AP/J2B (52.6 ◦C) ª I2AP/I2B
(52.2 ◦C) > H2AP/H2B (48.5 ◦C) (ª P1/N; 48.6 ◦C) ꢀ
K2AP/K2B (43.1 ◦C). Similarly, K6AP/K6B having six K–K pairs
◦
showed the lowest Tm (49.8 C), while those of H6AP/H6B and
I6AP/I6B were as high as 62.0 and 63.8 ◦C, respectively. This
is because the hydrophobic interactions of H–H pairs having
isopropylcyclohexane also efficiently worked even when pyrene
was inserted between them.
UV-Vis spectra of H2AP/H2B, H6AP/H6B, and P1/N are
shown in Fig. 7. Both H2AP/H2B and H6AP/H6B showed a
sharp band at 346 and 345 nm assignable to the pyrene moiety,
respectively. On the other hand, the absorption maximum of P1/N
was located at 351 nm, which was about 5 nm longer than that of
H2AP/H2B or H6AP/H6B and was broadened. These spectral
changes also support the shielding of pyrene from nucleobases
(vide infra).
3.2 Insulating ability of cyclohexyl base pairs
The incorporation of an H–H pair between nucleobases and
pyrene efficiently enhanced the emission of pyrene; the emission
intensity of H2AP/H2B was more than hundred times higher than
that of P1/N (see Fig. 5) because pyrene was efficiently shielded
from natural nucleobases.44 The suppressed interaction between
pyrene and nucleobases was also substantiated from a comparison
of the UV spectra of H2AP/H2B with P1/N, as shown in Fig.
7; both H2AP/H2B and H6AP/H6B showed a sharp band at
346 and 345 nm, respectively, whereas P1/N had a maximum at
351 nm with reduced absorbance (hypochromism). This difference
is interpreted as follows. In P1/N, pyrene was intercalated
between the adjacent nucleobases. Under these conditions, a
Discussion
3.1 High stability of multiple cyclohexyl base pairs
The incorporation of a single H–H pair slightly lowered duplex
stability, although cyclohexyl moieties were located inside the
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8313–8320 | 8317
©

View Article Online
strong excitonic interaction of pyrene with nucleobases induced
both bathochromicity and hypochromicity compared with isolated
pyrene absorption.45,46 However, in the cases of H2AP/H2B and
H6AP/H6B, isopropylcyclohexane moieties inserted between the
pyrene and nucleobases efficiently isolated pyrene and blocked
their excitonic interaction. Accordingly, the insertion of the H–H
pair induced hypochromicity and hyperchromicity.
interactions, so that together these pairs may have high mobility
even in the duplex. Accordingly, fixing an “insulator pair” by
incorporating associative interactions (such as hydrogen bonding)
would be the next step to suppress the undesired dynamics for
further development of the insulator pair.
4. Conclusion
Among the insulators synthesized in this study, J–J pairs
with biphenyl moieties did not show “shielding effects”, as
demonstrated by J2AP/J2B (Fig. 6). Since the J–J pair did
not destabilize the duplex, as shown in Table 1, it should be
located inside the DNA duplex.47,48 Indeed, bands of pyrene in
J1AP/J1B showed strong hypochromism compared to the other
duplexes (see Fig. S2 in the Supporting Information†), but note
that these spectral changes were not derived from the excitonic
interaction between the nucleobase and pyrene. The intercalated
biphenyl moiety that has a similar UV-Vis spectrum to the
nucleobase excitonically interacted with pyrene, resulting in strong
hypochromism. Since a J residue has two benzene rings, electron
transfer from nucleobases occurred through the p electrons of
the benzene rings and, thus, quenched their emission. From these
results, we think that at least one cyclohexane ring is necessary for
the enhancement of the quantum yield of pyrene.
The fluorescence intensity of I2AP/I2B was slightly lower than
that of H2AP/H2B, although its intensity was much higher than
P1/N. The same tendency was observed with I6AP/I6B and
H6AP/H6B. It is likely that the benzene ring of the I moiety
might slightly quench the fluorescence of pyrene. In contrast, K–
K pairs without benzene rings showed almost the same shielding
effect as H–H pairs, demonstrating that the isopropyl group in
the H moiety did not contribute to the shielding. However, the
absence of an isopropyl group destabilizes the duplex, which is a
severe problem when multiple chromophores and insulators are
introduced into DNA.41 Taken together, H–H pairs are the best
“insulator” among the four molecules.
We successfully prepared novel artificial bases that formed a
pair through hydrophobic interactions. These “base pairs” of
isopropylcyclohexane moieties showed even higher stability than
natural A–T pairs, although the moiety had no aromatic rings
or hydrogen bonds. Thermodynamic analyses demonstrated that
hydrophobic interactions between isopropyl groups contributed
significantly to the stability. In NOESY of duplexes containing
a cyclohexyl base pair (isopropylcyclohexane-methylcyclohexane
pair), distinct signals were observed between the methyl protons
of cyclohexyl moieties and imino protons of both the neighboring
base pairs. In addition, interhelical NOEs between cyclohexyl
moieties were also observed. CD spectra showed that a duplex
containing as many as six cyclohexyl base pairs still kept the native
B-form duplex. These results clearly demonstrated that cyclohexyl
moieties without aromatic rings formed a “base pair” inside the
DNA duplex.
When the artificial base pairs were introduced between pyrene
and nucleobases, cyclohexyl moieties worked as “insulators”, and
the emission intensity of pyrene greatly increased. The quantum
yield of a duplex containing one pair of isopropylcyclohexane
was as high as 0.19, which was more than hundred times higher
than the yield of a duplex without insulators. The cyclohexyl base
pairs were located between pyrene and the bases and efficiently
disturbed the electron transfer between them. Previously, we
proposed new methodology of accumulating fluorophores into
DNA duplex by mimicking quantum dots for its application
to “bright” labelling of biomolecule.41 We demonstrated that a
DNA-based fluorophore assembly could easily be labelled onto
long DNA only by treatment with ligase. However, available dyes
were limited to perylene derivatives since most of fluorophores
such as Cy3 significantly lowered the quantum yield by their
intercalation due to the strong quenching by natural nucleobases.
If we use cyclohexyl base pairs in place of natural base-pairs,
various fluorescent dyes can be accumulated without decreasing
quantum yield. Such fluorophore assembly has potential as
sensitive labelling agent of biomolecules to detect them at the
single-molecule level.
3.3 Effect of the number of insulators
In this study, the incorporation of a single H–H pair between
the pyrene and nucleobases drastically increased the emission
intensity. On the other hand, the introduction of two additional
pairs only moderately enhanced the emission of pyrene. In the
DNA duplex, electron transfer over short distances proceeds
by a superexchange mechanism.28–32 Consequently, the rate of
electron transfer should decay exponentially with the distance.
Here, the cyclohexyl insulator has no aromatic rings, so it could
not be involved in the electron transfer process. In our system, the
difference in quantum yield between H2AP/H2B and H6AP/H6B
is not large, whereas that between P1/N and H2AP/H2B is
massive. This high distance-dependence might reflect exponential
decay in the superexchange process. However, there are several
factors that might affect the apparent recovery of pyrene emission:
1) the hydrophobic environment created by cyclohexyl base pairs
might change the emission intensity of pyrene. In order to rule
out this possibility, a reporter dye monitoring the hydrophobicity
of the microenvironment should be introduced between the
cyclohexyl base pairs; 2) The mobility of the cyclohexane rings
might affect the efficiency. Note that neither H–H nor K–K
pairs were associated with hydrogen bonding or electrostatic
5. Experimental
5.1 Materials
All of the conventional phosphoramidite monomers, CPG
columns, reagents for DNA synthesis and Poly-Pak II cartridges
were purchased from Glen Research. Other reagents for the
synthesis of phosphoramidite monomers were purchased from
Tokyo Kasei Co., Ltd and Aldrich.
5.2 Synthesis of the DNA modified with pyrene and insulators
All of the modified ODNs were synthesized on an auto-
mated DNA synthesizer (ABI-3400 DNA synthesizer, Applied
8318 | Org. Biomol. Chem., 2011, 9, 8313–8320
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Biosystems) using phosphoramidite monomers bearing pyrene
and insulators. Phosphoramidite monomer tethering a methyl-
cyclohexane moiety (K residue) was synthesized as described
in the Supporting Information†. Phosphoramidite monomers of
other non-natural moieties and pyrene were prepared according
to previous reports.24,40 The coupling efficiency of the monomers
corresponding to the modified residues was as high as that of
the conventional monomers, as judged from the coloration of the
released trityl cation. After the recommended work-up, they were
purified by reversed phase (RP)-HPLC and were characterized by
MALDI-TOFMS (Autoflex, Bruker Daltonics). Purities of all the
ODNs are estimated over 99% from HPLC analysis.
to within 2.0 ◦C. The sample solutions were as follows: [NaCl] =
100 mM, pH 7.0 (10 mM phosphate buffer), [ODN] = 5.0 mM.
Thermodynamic parameters of duplexes (DH, DS) were de-
termined from 1/Tm versus ln(CT/4) plots by the following
equation: 1/Tm = R/DH ln(CT/4) + DS/DH, where CT is the
total concentration of ODNs. DG◦ was calculated from the DH
37
and DS values. The sample solutions were as follows: [NaCl] =
100 mM, pH 7.0 (10 mM phosphate buffer). The range of DNA
concentrations was 2–64 mM.
5.5 NMR measurements
NMR samples were prepared by dissolving three-times lyophilized
DNA in an H2O/D2O 9 : 1 solution containing 10 mM sodium
phosphate (pH 7.0) to give a duplex concentration of 1.7 mM.
NaCl was added to give a final sodium concentration of 200 mM.
NMR spectra were measured with a Varian INOVA spectrom-
eter (700 MHz) equipped for triple resonance at a probe temper-
ature of 275 K. Resonances were assigned by standard methods
using a combination of 1D, TOCSY (60 ms of mixing time), DQF-
COSY, and NOESY (150 ms of mixing time) experiments. All
spectra in the H2O/D2O 9 : 1 solution were recorded using the
3-9-19 WATERGATE pulse sequence for water suppression.
The MALDI-TOFMS data for the DNA were as follows:
H1A: Obsd 3962 (calcd for [H1A + H+]: 3964). H2A: Obsd 4283
(calcd for [H2A + H+]: 4283). H6A: Obsd 5558 (calcd for [H6A +
H+]: 5560). H1B: Obsd 3962 (calcd for [H1B + H+]: 3964). H2B:
Obsd 4281 (calcd for [H2B + H+]: 4283). H6B: Obsd 5559 (calcd
for [H6B + H+]: 5560). I1A: Obsd 3999 (calcd for [I1A + H+]: 3998).
I1B: Obsd 3999 (calcd for [I1B + H+]: 3998). I2B: Obsd 4351 (calcd
for [I2B + H+]: 4351). I6B: Obsd 5764 (calcd for [I6B + H+]: 5764).
J1A: Obsd 3992 (calcd for [J1A + H+]: 3992). J1B: Obsd 3992
(calcd for [J1B + H+]: 3992). J2B: Obsd 4341 (calcd for [J2B +
H+]: 4339). K1A: Obsd 3936 (calcd for [K1A + H+]: 3936). K1B:
Obsd 3937 (calcd for [K1B + H+]: 3936). K2A: Obsd 4225 (calcd
for [K2A + H+]: 4227). K2B: Obsd 4227 (calcd for [K2B + H+]:
4227). K6A: Obsd 5391 (calcd for [K6A + H+]: 5391). K6B: Obsd
5393 (calcd for [K6B + H+]: 5391). P1: Obsd 4081 (calcd for [P1 +
H+]: 4082). H1C: Obsd 2112 (calcd for [H1C + H+]: 2112). K1D:
Obsd 2084 (calcd for [K1D+H+]: 2084). H2AP: Obsd 4719 (calcd
for [H2AP + H+]: 4720). I2AP: Obsd 4788 (calcd for [I2AP + H+]:
4788). J2AP: Obsd 4776 (calcd for [J2AP + H+]: 4776). K2AP:
Obsd 4664 (calcd for [K2AP+H+]: 4664). H6AP: Obsd 5995 (calcd
for [H6AP + H+]: 5997). I6AP: Obsd 6201 (calcd for [I6AP + H+]:
6201). K6AP: Obsd 5826 (calcd for [K6AP + H+]: 5829).
5.6 Computer modeling
Molecular modeling by conformational energy minimization
was performed with Insight II/Discover 3 software (Molecular
Simulation, Inc.) on a Silicon Graphics Octane workstation with
the operating system IRIX64 Release 6.5 and AMBER was used
for the calculations. The results of the NMR analyses served as a
starting point for the modeling (no restraints were used for energy
minimization).
Acknowledgements
This work was partially supported by a Grant-in-Aid for Scientific
Research (A) (21241031) and a Grant-in-Aid for Young Scientists
(B) (22750149) from the Ministry of Education, Culture, Sports,
Science and Technology, Japan. Partial support by SENTAN
program, Japan Science and Technology Agency (JST), is also
acknowledged.
5.3 Spectroscopic measurements
Fluorescence spectra were measured on a JASCO model FP-
6500 with a microcell. The excitation wavelength was 345 nm.
The sample solutions were as follows: [NaCl] = 100 mM, pH 7.0
(10 mM phosphate buffer), [ODN] = 1.0 mM. Quantum yields
were determined from the quantum yield of pyrene in N2-bubbled
cyclohexane (0.65) as a reference.
UV-Vis and CD spectra were measured on a Shimadzu UV-1800
and a JASCO model J-820, respectively, with a 10-mm quartz cell
equipped with programmed temperature controllers. The sample
solutions were as follows: [NaCl] = 100 mM, pH 7.0 (10 mM
phosphate buffer), [ODN] = 5.0 mM.
Notes and references
1 B. A. Schweitzer and E. T. Kool, J. Am. Chem. Soc., 1995, 117, 1863–
1872.
2 C. Brotschi, A. Ha¨berli and C. J. Leumann, Angew. Chem., Int. Ed.,
2001, 40, 3012–3014.
3 H. Liu, J. Gao, S. R. Lynch, Y. D. Saito, L. Maynard and E. T. Kool,
Science, 2003, 302, 868–871.
4 N. Minakawa, N. Kojima, S. Hikishima, T. Sasaki, A. Kiyosue, N.
Atsumi, Y. Ueno and A. Matsuda, J. Am. Chem. Soc., 2003, 125, 9970–
9982.
5.4 Measurement of the melting temperature
5 S. Matsuda and F. E. Romesberg, J. Am. Chem. Soc., 2004, 126, 14419–
The melting curve of duplex DNA was obtained with a Shimadzu
UV-1800 by measurement of the change in absorbance at 260 nm
versus temperature. The melting temperature (Tm) was calculated
from the maximum in the first derivative of the melting curve. Both
the heating and the cooling profiles were measured and it◦s average
was determined as Tm. The temperature ramp was 0.5 C min-1
and the Tms determined from heating and cooling profiles agreed
14427.
6 N. Amann, R. Huber and H.-A. Wagenknecht, Angew. Chem., Int. Ed.,
2004, 43, 1845–1847.
7 Z. Yang, D. Hutter, P. Sheng, A. M. Sismour and S. A. Benner, Nucleic
Acids Res., 2006, 34, 6095–6101.
8 I. Hirao, T. Mitsui, M. Kimoto and S. Yokoyama, J. Am. Chem. Soc.,
2007, 129, 15549–15555.
9 S. Hainke and O. Seitz, Angew. Chem., Int. Ed., 2009, 48, 8250–8253.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8313–8320 | 8319
©

View Article Online
10 F. D. Lewis, T. Wu, E. L. Burch, D. M. Bassani, J.-S. Yang, S. Schneider,
W. Jaeger and R. L. Letsinger, J. Am. Chem. Soc., 1995, 117, 8785–8792.
11 F. D. Lewis, Y. Zhang and R. L. Letsinger, J. Am. Chem. Soc., 1997,
119, 5451–5452.
12 S. M. Langenegger and R. Ha¨ner, Helv. Chim. Acta, 2002, 85, 3414–
3421.
33 M. Manoharan, K. L. Tivel, M. Zhao, K. Nafisi and T. L. Netzel, J.
Phys. Chem., 1995, 99, 17461–17472.
34 P. L. Paris, J. M. Langenhan and E. T. Kool, Nucleic Acids Res., 1998,
26, 3789–3793.
35 K. Yamana, T. Iwai, Y. Ohtani, S. Sato, M. Nakamura and H. Nakano,
Bioconjugate Chem., 2002, 13, 1266–1273.
36 K. Fujimoto, H. Shimizu and M. Inouye, J. Org. Chem., 2004, 69,
3271–3275.
13 U. B. Christensen and E. B. Pedersen, Helv. Chim. Acta, 2003, 86,
2090–2097.
14 U. B. Christensen and E. B. Pedersen, Nucleic Acids Res., 2002, 30,
4918–4925.
15 S. M. Langenegger and R. Ha¨ner, Chem. Commun., 2004, 2792–2793.
16 P. J. Hrdlicka, B. R. Babu, M. D. Sorensen, N. Harrit and J. Wengel, J.
Am. Chem. Soc., 2005, 127, 13293–13299.
17 R. Ha¨ner, F. Garo, D. Wenger and V. L. Malinovskii, J. Am. Chem.
Soc., 2010, 132, 7466–7471.
37 A. Okamoto, K. Kanatani and I. Saito, J. Am. Chem. Soc., 2004, 126,
4820–4827.
38 M. Raunkjær, K. F. Haselmann and J. Wengel, J. Carbohydr. Chem.,
2005, 24, 475–502.
39 H. Kashida, H. Asanuma and M. Komiyama, Chem. Commun., 2006,
2768–2770.
40 H. Kashida, T. Takatsu, K. Sekiguchi and H. Asanuma, Chem.–Eur.
J., 2010, 16, 2479–2486.
41 H. Kashida, K. Sekiguchi, X. Liang and H. Asanuma, J. Am. Chem.
Soc., 2010, 132, 6223–6230.
42 J. N. Wilson, Y. Cho, S. Tan, A. Cuppoletti and E. T. Kool,
ChemBioChem, 2008, 9, 279–285.
43 This quantum yield is lower than pyrene-modified DNA in other
reports. In this study, 1-pyrenebutyric acid unit was introduced
into DNA through D-threoninol. Accordingly, pyrene is intercalated
between base pairs and is highly quenched by nucleobases. See refs 35
and 38.
44 We could not calculate the increasing rate of emission intensity of
H2AP/H2B directly because the emission intensity of P1/N was much
lower than that of H2AP/H2B. The emission of P1/H2B at high
sensitivity was about 36 times higher than that of P1/N. In addition,
the emission intensity of H2AP/H2B at medium sensitivity was about
4.5 times higher than that of P1/H2B. See Fig. S3 in the Supporting
Information† for actual data.
45 A. Mahara, R. Iwase, T. Sakamoto, T. Yamaoka, K. Yamana and A.
Murakami, Bioorg. Med. Chem., 2003, 11, 2783–2790.
46 A. Okamoto, K. Kanatani, Y. Ochi, Y. Saito and I. Saito, Tetrahedron
Lett., 2004, 45, 6059–6062.
18 S. A. Benner, Acc. Chem. Res., 2004, 37, 784–797.
19 K. Tanaka and M. Shionoya, Chem. Lett., 2006, 35, 694–699.
20 A. T. Krueger and E. T. Kool, Curr. Opin. Chem. Biol., 2007, 11, 588–
594.
ˇ
21 J. Stambasky´, M. Hocek and P. Kocˇovsky´, Chem. Rev., 2009, 109,
6729–6764.
22 V. L. Malinovskii, D. Wenger and R. Ha¨ner, Chem. Soc. Rev., 2010, 39,
410–422.
23 M. Kaufmann, M. Gisler and C. J. Leumann, Angew. Chem., Int. Ed.,
2009, 48, 3810–3813.
24 H. Kashida, K. Sekiguchi and H. Asanuma, Chem.–Eur. J., 2010, 16,
11554–11557.
25 H. Kashida, X. G. Liang and H. Asanuma, Curr. Org. Chem., 2009, 13,
1065–1084.
26 M. Liu, X.-A. Mao, C. Ye, H. Huang, J. K. Nicholson and J. C. Lindon,
J. Magn. Reson., 1998, 132, 125–129.
27 H. Kashida, H. Ito, T. Fujii, T. Hayashi and H. Asanuma, J. Am. Chem.
Soc., 2009, 131, 9928–9930.
28 G. B. Schuster, Acc. Chem. Res., 2000, 33, 253–260.
29 F. D. Lewis, R. L. Letsinger and M. R. Wasielewski, Acc. Chem. Res.,
2000, 34, 159–170.
30 B. Giese, Annu. Rev. Biochem., 2002, 71, 51–70.
31 H.-A. Wagenknecht, Nat. Prod. Rep., 2006, 23, 973–1006.
32 J. C. Genereux and J. K. Barton, Chem. Rev., 2010, 110, 1642–1662.
47 Z. Johar, A. Zahn, C. J. Leumann and B. Jaun, Chem.–Eur. J., 2008,
14, 1080–1086.
48 A. Zahn and C. J. Leumann, Chem.–Eur. J., 2008, 14, 1087–1094.
8320 | Org. Biomol. Chem., 2011, 9, 8313–8320
This journal is
The Royal Society of Chemistry 2011
©