Naphthyridine-Benzoazaquinolone: Evaluation of a Tricyclic System for the Binding to (CAG)n Repeat DNA and RNA

Article abstract of DOI:10.1002/asia.201600527

The expansion of CAG repeats in the human genome causes the neurological disorder Huntington's disease. The small-molecule naphthyridine-azaquinolone NA we reported earlier bound to the CAG/CAG motif in the hairpin structure of the CAG repeat DNA. In order to investigate and improve NA-binding to the CAG repeat DNA and RNA, we conducted systematic structure-binding studies of NA to CAG repeats. Among the five new NA derivatives we synthesized, surface plasmon resonance (SPR) assay showed that all of the derivatives modified from amide linkages in NA to a carbamate linkage failed to bind to CAG repeat DNA and RNA. One derivative, NBzA, modified by incorporating an additional ring to the azaquinolone was found to bind to both d(CAG)₉ and r(CAG)₉. NBzA binding to d(CAG)₉ was similar to NA binding in terms of large changes in the SPR assay and circular dichroism (CD) as well as pairwise binding, as assessed by electron spray ionization time-of-flight (ESI-TOF) mass spectrometry. For the binding to r(CAG)₉, both NA and NBzA showed stepwise binding in ESI-TOF MS, and NBzA-binding to r(CAG)₉ induced more extensive conformational change than NA-binding. The tricyclic system in NBzA did not show significant effects on the binding, selectivity, and translation, but provides a large chemical space for further modification to gain higher affinity and selectivity. These studies revealed that the linker structure in NA and NBzA was suitable for the binding to CAG DNA and RNA, and that the tricyclic benzoazaquinolone did not interfere with the binding.

Full text of DOI:10.1002/asia.201600527

DOI: 10.1002/asia.201600527
Full Paper
Binding to Trinucleotide Repeats
Naphthyridine-Benzoazaquinolone: Evaluation of a Tricyclic
System for the Binding to (CAG)n Repeat DNA and RNA
Jinxing Li +,[a] Akihiro Sakata+,[a] Hanping He,[b] Li-Ping Bai,[c] Asako Murata,[a]
Chikara Dohno,[a] and Kazuhiko Nakatani*[a]
Abstract: The expansion of CAG repeats in the human
genome causes the neurological disorder Huntington’s dis-
ease. The small-molecule naphthyridine-azaquinolone NA we
reported earlier bound to the CAG/CAG motif in the hairpin
structure of the CAG repeat DNA. In order to investigate and
improve NA-binding to the CAG repeat DNA and RNA, we
conducted systematic structure-binding studies of NA to
CAG repeats. Among the five new NA derivatives we synthe-
sized, surface plasmon resonance (SPR) assay showed that
all of the derivatives modified from amide linkages in NA to
a carbamate linkage failed to bind to CAG repeat DNA and
RNA. One derivative, NBzA, modified by incorporating an ad-
ditional ring to the azaquinolone was found to bind to both
d(CAG)9 and r(CAG)9. NBzA binding to d(CAG)9 was similar to
NA binding in terms of large changes in the SPR assay and
circular dichroism (CD) as well as pairwise binding, as as-
sessed by electron spray ionization time-of-flight (ESI-TOF)
mass spectrometry. For the binding to r(CAG)9, both NA and
NBzA showed stepwise binding in ESI-TOF MS, and NBzA-
binding to r(CAG)9 induced more extensive conformational
change than NA-binding. The tricyclic system in NBzA did
not show significant effects on the binding, selectivity, and
translation, but provides a large chemical space for further
modification to gain higher affinity and selectivity. These
studies revealed that the linker structure in NA and NBzA
was suitable for the binding to CAG DNA and RNA, and that
the tricyclic benzoazaquinolone did not interfere with the
binding.
tients with myotonic dystrophy type 1.[7–9] The CGG repeat ex-
pansion causes Fragile X syndrome upon expansion of up to
more than 200 repeats.[10,11] These TNR sequences share the
common sequence motif of CXG, where X is A, T, C, and G,
while there are other types of TNR such as GAA, which causes
Friedreich’s ataxia upon expansion in the FXN gene.[12,13]
Introduction
The expansion of trinucleotide repeat (TNR) sequences in the
human genome is known as causative of more than 40 heredi-
tary neurological disorders.[1–4] The CAG sequence in the
coding region of the protein huntingtin is 4–100 repeats for
patients with Huntington’s disease, whereas that is about 6–
36 repeats for healthy individuals.[5,6] Aberrant and enormous
expansion of the CTG sequence of 5–38 repeats in the 3’-UTR
of the DMPK gene up to 40–1000 repeats was observed for pa-
The mechanisms of dynamic instability of TNR sequence in-
volving repeat expansion and contraction depend on the type
of TNR sequences, the involvement of repair systems, and the
type of cells and tissues.[14–17] The common feature of the
mechanism in TNR expansion and contraction, however, in-
volves non-canonical secondary structures produced on the
TNR sequences.[18–23] Due to C-G base-pairing, the expanded
CXG repeats can fold into a hairpin secondary structures con-
taining a number of the palindromic CXG/CXG motif, which
[a] Dr. J. Li ,+ A. Sakata,+ Prof. Dr. A. Murata, Prof. Dr. C. Dohno,
Prof. Dr. K. Nakatani
Department of Regulatory Bioorganic Chemistry
The Institute of Scientific and Industrial Research
Osaka University
8-1 Mihogaoka, Ibaraki 567-0047 (Japan)
E-mail: nakatani@sanken.osaka-u.ac.jp
[b] Prof. Dr. H. He
Chemistry and Chemical Engineering College
Hubei University
Road Youyi 368, Wuchang, Wuhan, Hubei 430062 (China)
[c] Prof. Dr. L.-P. Bai
State Key Laboratory of Quality Research in Chinese Medicine, and Macau
Institute for Applied Research in Medicine and Health
Macau University of Science and Technology
Avenida Wai Long, Taipa, Macau (China)
Figure 1. (a) A possible hairpin secondary structure of (CXG)n, (b) the struc-
ture of naphthyridine-azaquinolone (NA), and (c) 2:1 binding mode of NA to
a 5’-d(CAG)-3’/5’-d(CAG)-3’ motif. The red and blue blocks represent naph-
thyridine and azaquinolone units, respectively.
[+] These authors contributed equally to these work.
Supporting information for this article can be found under http://
dx.doi.org/10.1002/asia.201600527.
Chem. Asian J. 2016, 00, 0 – 0
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
hold a X-X mismatch flanked by two C-G base pairs (Figure 1a).
One of the mechanisms of TNR expansion is believed to in-
volve the polymerase stall during the replication of TNR se-
quence and subsequent slippage of the TNR sequence to form
hairpin structures on the synthesizing strand.[22,23] Since the
stability of the hairpin secondary structures involving CXG/CXG
motifs increases as the repeat length increases, the expanded
CXG sequence is more susceptible to the formation of the hair-
pin structure and more likely to undergo further expan-
sion.[24–26]
Results
Design and synthesis of NA derivatives
NA consisted of two units of aromatic heterocycles, which
were connected by the linker through an amide linkage. This
linker was adopted from the first-generation molecule of our
mismatch-binding ligands naphthyridine dimer, which selec-
tively bound to the G-G mismatch in double-stranded
DNA.[34–36] The amide linkage to the 2-amino group of 1,8-
naphthyridine was rather weak due to the electron-withdraw-
ing property of 1,8-naphthyridine and, therefore, underwent
unfavorable hydrolysis at a high temperature even in a neutral
pH. To circumvent this potential instability, the amide linkage
was replaced by a carbamate linkage for connecting the amino
group of 1,8-naphthyridine. A new G-G mismatch binding mol-
ecule having a carbamate linkage showed an improved chemi-
cal stability and affinity,[29,37] which is likely due to the in-
creased p-stacking at the carbamate moiety and the longer
linker length by two atoms than the amide linkage. According
to the successful transformation in G-G mismatch-binding mol-
ecules, the amide linkage connecting heterocycle units in NA
was modified to the carbamate linkage.
We have discovered small molecules that selectively bind to
the CXG/CXG motifs and promote the formation of hairpin sec-
ondary structures on the CXG repeat sequences.[27–30] Naph-
thyridine-azaquinolone NA is the first discovered molecule that
binds to the CAG/CAG motif produced in the hairpin secon-
dary structure of the CAG repeat DNA sequence (Figure 1b).[27]
The NA-bound structure on the CAG/CAG motif was deter-
mined by NMR spectroscopy to have the following structural
characteristics (Figure 1c): 1) the naphthyridine unit in NA
bound to the guanine by three hydrogen bonds, 2) cytosine
that base-paired to the guanine was forced to flip out from
the p-stack and had little effect on the stability of the complex,
3) the azaquinolone unit in NA bound to the adenine by two
hydrogen bonds, 4) two molecules of NA simultaneously
bound to the CAG/CAG motif, and 5) naphthyridine and aza-
quinolone units in a different molecule were well stacked with
each other and neighboring base pairs. In addition to the
d(CAG)n repeat, r(CAG)n repeat, a transcript of the expanded
d(CAG)n sequence has drawn significant recent scientific inter-
ests because the expanded r(CAG)n repeat was found to be
translated without the presence of the AUG start codon.[31–33]
This phenomenon is called repeat associated non-ATG transla-
tion (RAN translation) and is supposed to involve the non-can-
onical hairpin structure of the expanded r(CAG)n repeat. Small
molecules binding to the hairpin secondary structure of the
CAG repeat RNA with high affinity hold potentials to modulate
the repeat expansion and contraction, and RAN translation.
To investigate the binding of NA to the r(CAG)n repeat and
further improvements of the NA structure regarding the affini-
ty and selectivity not only to d(CAG)n but also to r(CAG)n re-
peats, we conducted systematic structure-binding studies of
NA on the binding to these CAG repeat sequences. Here we
describe the effect of the structural modification of the linker
connecting naphthyridine and azaquinolone units and the ex-
panded aromatic system from bicyclic azaquinolone to the tri-
cyclic benzoazaquinolone on the binding to d(CAG)9 and
r(CAG)9 investigated by SPR assay, CD spectra, and ESI-TOF MS.
These studies revealed that, among five derivatives of NA, the
ring-expanded derivative NBzA with the amide linkage bound
to d(CAG)n repeat in a pairwise manner similar to NA. In con-
trast, NA and NBzA showed the binding to r(CAG)n repeat with
decreased affinity, and the binding was not pairwise but in
a stepwise manner. The change in CD spectra of r(CAG)9 was
more extensive for NBzA than NA, suggesting the effect of tri-
cyclic system of NBzA. These observations for NBzA binding
are important clues for improving the affinity and selectivity of
ligands targeting the genetically important r(CAG)n repeat.
For the systematic structure-binding studies, we separated
NA into two parts in design, namely, amide-linked naphthyri-
dine N and amide-linked azaquinolone A (Figure 2). The
amide-linked 1,8-naphthyridine N was modified to the carba-
mate-linked 1,8-naphthyridine NC, which might have a better
Figure 2. Structures of NA derivatives discussed in these studies. The combi-
nation of G-binding units (N and NC) shown in the left with A-binding units
(A, BzA, AC, and BzAC) shown in the right provided six derivatives: NA,
NBzA, NAC, NBzAC, NCA, and NCAC.
stacking ability with the neighboring bases and flexibility in
the linker. With regard to the 8-azaquinolone part, the amide-
linked 8-azaquinolone A was also changed to the carbamate-
linked 8-azaquinolone AC. For these structural changes in the
linker moiety, however, we could not keep the linker length
same as that of parent NA, because of the instability of hemia-
minal toward the acid hydrolysis and the facile cyclization into
cyclic carbamate[37] (Figure 3).
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
repeat folds into a hairpin secondary structure. The number of
potential binding sites, however, could not be explicitly proven
due to the possibility of forming diverse hairpin structures. Fur-
thermore, the sensor surface with hairpin secondary structures
of CXG repeat DNA and RNA may not reproducible upon dena-
turation and renaturation processes. With these ambiguities in
the binding stoichiometry and reproducibility of the surface,
the SPR sensorgrams obtained for ligand binding to CXG
repeat sensor were discussed only in qualitative aspects.
Figure 3. Limitation in the linker structures due to (a) instability of the hemi-
aminal structure and (b) facile intramolecular nucleophilic attack of nitrogen
to a carbamate carbonyl group.
In addition, the 8-azaquinolone part was further modified by
introducing a third ring system in benzoazaquinolone BzA. In
the NA-CAG/CAG complex, NMR studies clearly revealed that
8-azaquinolone bound to the adenine through two hydrogen
bonds mimicking the T-A base pairing.[27] However, the azaqui-
nolone-A binding was likely an additional binding concomitant
with the strong 1,8-naphthyridine-G binding based
on the previous structure–activity studies on the NA
binding to the G-A mismatch.[38] In fact, a naphthyri-
dine dimer, where the 8-azaquinolone part of NA was
substituted with 1,8-naphthyridine, still showed the
affinity to the G-A mismatch, although the affinity
was not as high as that of NA. To enhance the bind-
ing of the 8-azaquinolone part and, eventually, in-
crease the affinity of the molecule to the CAG/CAG
motif, we investigated the effect of the third ring
system that may provide a better stacking possibility
with the neighboring bases. The BzA moiety was also
transformed into the carbamate-linked unit BzAC.
With these basic designs in mind, combination of the
1,8-naphthyridine parts (N and NC) with the 8-azaqui-
nolone parts (A, BzA, AC, and BzAC) yielded five new
molecules, NBzA, NAC, NBzAC, NCA, and NCAC. The
synthesis of these molecules is described in the Ex-
perimental Section. In brief, introduction of a carba-
mate linkage was accomplished by using N-Boc-((3-
hydroxypropyl)amino)propanoic acid (for NCA, NAC,
and NBzAC) and N-Boc-bis(3-hydroxypropyl)amine
(for NCAC and NCAC). The benzoazaquinolone unit
was synthesized from 2-amino-6-methylpyridine by
way of 3-methylbenzo[c][1,8]naphthyridin-6(5H)-one
and subsequent transformation of the methyl group.
SPR sensorgrams of ligand binding to d(CAG)9-immobilized
surface showed strong responses for NA and NBzA with the
response unit reaching about 140 and 225 response units (RU),
respectively, when both ligands were applied at 2 mm concen-
tration. (Figure 4) In contrast, the other four ligands NCA, NAC,
NCAC, and NBzAC showed a much decreased response below
Binding analysis of NA derivatives to (CAG)n repeat
DNA by SPR assay
At first, we investigated the binding of NA and five
NA derivatives to d(CAG)9 by the SPR assay using
d(CAG)9-immobilized surface prepared on a streptavi-
din-coated sensor (SA chip, GE Healthcare). In gener-
al, the analysis of SPR sensorgrams provides with ki-
netic parameters for the binding. This requires fitting
the sensorgrams to the theoretical binding isotherm.
When the host has a single binding site for the
Figure 4. SPR analysis of ligand binding to the d(CAG)9-immobilized surface. Each ligand
was added stepwise at 0 (black), 0.13 (green), 0.25 (pink), 0.5 (red), 1.0 (pale blue), and
2.0 mm (blue). The sensors were exposed to the ligand for 0–60 s. d(CAG)9 was immobi-
lized on the SA sensor chip for 497 RU. (a) NA, (b) NAC, (c) NCA, (d) NCAC, (e) NBzA, and
ligand, we could use the binding stoichiometry of
1:1. In the CXG repeat DNA and RNA, we speculated
the potential binding site as the X-X mismatch
flanked by C-G base pairs (CXG/CXG) when the CXG (f) NBzAC.
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
10 RU even when 2 mm ligand was applied. Further-
more, the shape of response curves of these weak
binders showing fast association and dissociation
was totally different from those observed for NA and
NBzA. The large responses observed for NA and
NBzA with slow association and dissociation showed
the strong binding of these ligands to the d(CAG)9
repeat. These SPR data for NA were fully consistent
with those we reported before.[34] The similarity in
the SPR response curves obtained for NBzA and NA
suggested that NBzA binding to d(CAG)9 shares the
characteristics of mode of the NA binding to d(CAG)9.
Binding analysis of NA derivatives to (CAG)n repeat
RNA by SPR assay
Then, we investigated the binding of NA and five NA
derivatives to the r(CAG)9 repeat by the SPR assay
using r(CAG)9-immobilized surface prepared on the
SA chip (GE healthcare). SPR sensorgrams for the
binding of six ligands to r(CAG)9-immobilized surface
showed a marked ligand dependence (Figure 5).
Among six ligands, the strongest response of ca. 13
RU was observed for NBzA when 2 mm of the ligand
was used. NA also showed a response to the r(CAG)9
repeat with lower efficiency. NA and NBzA showed
a concentration-dependent SPR response, showing
that NA and NBzA bound to the r(CAG)9. All four li-
gands NCA, NAC, NCAC, and NBzAC did not show
significant binding in the SPR analyses to r(CAG)9 at
the concentration below 2 mm, while these ligands
showed SPR responses when the concentration in-
creased up to 50 mm of ligand (data not shown).
These data indicated that four ligands should have
a much lower affinity to r(CAG)9 than that of NA and
NBzA.
We also conducted single cycle SPR analysis for NA
and NBzA binding to d(CAG)n and r(CAG)n, confirm-
ing their binding (Figure 6 and 7). According to the
shape of response curves, both association to and
dissociation from d(CAG)9 seems slower for NA than
NBzA. Finally, we investigated if NA and NBzA might
bind to other trinucleotide repeat sequences. Single
Figure 5. SPR analysis of ligand binding to the r(CAG)9-immobilized surface. Each ligand
was added stepwise at 0 (black), 0.13 (green), 0.25 (pink), 0.5 (red), 1.0 (pale blue), and
2.0 mm (blue). The sensors were exposed to the ligand for 0-60 s. r(CAG)9 was immobi-
lized on the SA sensor chip for 524 RU. (a) NA, (b) NAC, (c) NCA, (d) NCAC, (e) NBzA, and
(f) NBzAC.
cycle kinetic analysis firmly confirmed that NA- and NBzA-
binding to r(CCG)9, r(CGG)9, and r(CUG)9 were not significant
up to 3 mm ligand concentration (Figures S1 and S2).
became negative in addition to the appearance of a large neg-
ative induced CD at 320 nm. NBzA-biding to d(CAG)9 also pro-
duced a negative CD signal at 270 nm and an induced signal
at 320 nm. (Figure 8c) The positive peak observed at 250 nm
for NA-binding to d(CAG)9 shifted to 235 nm in the NBzA-
d(CAG) complex probably because of a large absorption band
of NBzA at 230 nm.
The CD spectra of r(CAG)9 showed an A-form like structure
in the absence of ligands (Figure 8b). A negative peak was ob-
served at 240 nm and a positive peak was at 265 nm. Upon ad-
dition of NA, both the positive and negative peaks shifted
toward a short wavelength by 5 to 10 nm with concomitant
appearance of broad negative induced CD peak at 320 nm. All
the CD peaks were rather broad. In contrast, the addition of
CD spectra change on ligand binding
To gain information into the secondary structure, we measured
the CD spectra of d(CAG)9 and r(CAG)9 in the presence and ab-
sence of NA and NBzA. As we reported, the CD spectra of
d(CAG)9 changed markedly upon binding with NA[27] (Fig-
ure 8a). Thus, d(CAG)9 showed CD spectra similar to that of B-
form dsDNA with a negative peak at 250 nm and a positive
peak at 270 nm. Upon binding with NA, the negative peak at
250 nm became a positive peak and the peak at 270 nm
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 6. Single-cycle kinetic analysis of the ligands to d(CAG)9 (497 RU) by
SPR assay. The ligand (a) NA or (b) NBzA was sequentially added at 0.75, 1.0,
1.5, 2.0, and 3.0 mm.
Figure 7. Single-cycle kinetic analysis of the ligands to r(CAG)9 (524 RU) by
SPR assay. The ligand (a) NA or (b) NBzA was sequentially added at 0.75, 1.0,
1.5, 2.0, and 3.0 mm.
NBzA increased the molar ellipticity of the positive peak at
280 nm with a shift toward a longer wavelength by 15 nm
(Figure 8d). Since the absorption at 280 nm was very low for
NA and NBzA (Figure S3), the strong positive CD peak at
280 nm in NBzA-bound r(CAG)9 suggested the structural
change on RNA. The induced CD signals at 320 nm also
became much more apparent than that of NA-bound complex.
In addition to the major induced CD peak, two weak but sig-
nificant peaks at 345 nm (negative) and 355 nm (positive) were
clearly seen for NBzA. By comparing the CD spectra of NA-
and NBzA-bound r(CAG)9 repeat, it is likely that the binding of
NBzA induced more extensive conformational change on RNA.
obtained for the complex with a A-A mismatch flanking two C-
G base pairs, a basic unit of CAG/CAG triad produced in the
hairpin secondary structure of d(CAG)n repeat (cf. Figure 1).
ESI-TOF MS measurements of d(CAG)n repeat with NA suc-
cessfully reproduced the molecular ions containing an even
number of NA molecules as shown in Figure 9b. Molecular
ions containing an odd number of NA molecules were almost
negligible in terms of intensity. Similar mass spectra showing
molecular ions containing an even number of NBzA molecules
were obtained for NBzA (Figure 9c), suggesting that the mode
of NBzA binding to d(CAG)n repeat is most likely involved in
a pairwise binding as we observed for NA.
In marked contrast, the MS results for r(CAG)n repeat are to-
tally different from the ESI-TOF MS of d(CAG)n repeat in terms
of the number of ligand molecules included in the observed
molecular ions. Thus, r(CAG)9 showed 5- and 6-ions in the MS
(Figure 10a). These ions decreased the intensity in the pres-
ence of NA with concomitant appearance of new ions corre-
sponding to 5- and 6-ions of r(CAG)9 +1NA and r(CAG)9 +2NA
complexes (Figure 10b). In the presence of NBzA, ions corre-
sponding to complexes containing 1 to 3 NBzA were observed
as well (Figure 10c). However, the intensity of ions correspond-
ing to the complex containing 2 NBzA molecules was much
stronger in the intensity than those containing 1 and 3 NBzA
molecules. With an increase of NBzA concentration, ions of the
Stoichiometry of NA- and NBzA-binding to d(CAG)n and
r(CAG)n repeats
On the basis of SPR and CD measurements, the binding of NA
and NBzA to r(CAG)n repeat is suggested to be significantly dif-
ferent from that to d(CAG)n in terms of the mode of binding.
To gain further insight into the mode of the binding of the
two ligands to d(CAG)n and r(CAG)n repeat, ESI-TOF MS was
carried out for these ligands and repeats. In our previous stud-
ies on NA, we have revealed that NA binds to d(CAG)n repeat
in a pairwise manner, showing intense molecular ions contain-
ing an even number of NA molecules.[27] This pairwise binding
of NA to d(CAG)n repeat was supported by the NMR structure
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
complexes containing 4 and 5 NBzA were also ob-
served (data not shown), suggesting that the prefer-
ence of an even number of ligands in the binding to
d(CAG)n repeat was not likely applicable to NA- and
NBzA-binding to r(CAG)9 repeat. On the basis of
these observations, we concluded that the mode of
NA and NBzA binding to r(CAG)9 does not involve
a pairwise binding.
The effect of an additional ring system in NBzA in
the binding to r(CAG)n repeat
For the binding to r(CAG)9, NBzA induced more
prominent structural changes upon binding as
shown in CD spectra and the NBzA-bound complex
showed better stability in ESI-TOF MS than the NA-
bound complex. Single-cycle kinetic analysis by SPR
suggested faster association and dissociation kinetics
for NBzA than NA. These different binding character-
istics of NBzA from those of NA are obviously due to
the tricyclic benzoazaquinolone system implemented
into the structure. The benzoazaquinolone in NBzA is
anticipated to have a better stacking interaction with
neighboring bases than the parent azaquinolone.
Ding and Ellestad studied the effect of the hydro-
phobicity of the molecule in the binding to DNA by
changing the counter anion of sodium cation from
Figure 8. CD spectra of d(CAG)9 and r(CAG)9 (3 mm) in the presence of NA and NBzA
(50 mm) in 10 mm sodium cacodylate buffer (pH 7.0) and 100 mm NaCl at room tempera-
ture. (a) NA-d(CAG)9, (b) NA-r(CAG)9, (c) NBzA-d(CAG)9, and (d) NBzA-r(CAG)9
À
a weakly hydrated ClO4 anion to a strongly hydrated
Figure 9. ESI-TOF-MS analysis of (a) d(CAG)9 (2 mm), (b) d(CAG)9 with NA
(10 mm), and (c) d(CAG)9 with NBzA (10 mm).
Figure 10. ESI-TOF-MS analysis of (a) r(CAG)9 (2 mm), (b) r(CAG)9 with NA
(10 mm), and (c) r(CAG)9 with NBzA (10 mm).
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
2À
2À
SO4 anion.[39] The SO4 anion showed “salting out” effect by
ed by a reporter system involving Renilla luciferase (Rluc) and
firefly luciferase (Fluc)[41,42] (Figure 12). We have constructed
a vector containing d(CAG)89 repeat between Rluc and Fluc
genes and a control vector that did not contain the repeat
(Figure 12A and 12B). After in vitro transcription, the RNA was
in vitro translated with Rabbit Reticulocyte Lysate System in
the absence (w/o) or in the presence of NA or NBzA in differ-
ent concentrations. The activities of Rluc and Fluc were mea-
sured by the luciferase assay kit and compared with those ob-
tained by in vitro translation of control vector. For the transla-
tion of luciferases on the control RNA (Figure 12B), both NA
and NBzA showed a tendency of suppressing the translation
of both luciferases as the concentration of ligands increased.
The suppression rate of the control RNA was 0.53 and 0.41 for
Rluc and Fluc, respectively, (Figure 12B-a and B-c) at 100 mm of
NA, giving the ratio of Fluc and Rluc activities of 0.78 (Fig-
ure 12B-e). NBzA also showed the similar effects on translation
of luciferases on control RNA, providing 0.41 and 0.30 for Rluc
and Fluc, respectively, (Figure 12B-b and B-d) at 100 mm. The
ratio of two luciferase activities was 0.71. Comparing NA and
NBzA, the translation suppression was more significant for
NBzA (cf. B-a and B-b, and B-c and B-d).
À
increasing the hydrophobic interaction, whereas the ClO4
anion showed the opposite effect.[40] The ClÀ anion had inter-
mediate effect between these two anions. The CD spectra of
r(CAG)9 (3 mm) in the presence of NA and NBzA (50 mm) were
measured in the presence of 100 mm of NaClO4, NaCl, and
Na2SO4. While the CD spectra of r(CAG)9 was not affected by
the anion at all (Figure S4), significant differences were ob-
served in the presence of NA (Figure 11a). The CD bands in-
For the translation of luciferases containing (CAG)89 repeat
between Rluc and Fluc (Figure 12A), translation suppression of
Fluc by NA and NBzA was somewhat less efficient as com-
pared to those observed for the translation of Fluc on the con-
trol RNA. Thus, the activity of Fluc translated from repeat con-
taining RNA was 0.50 and 0.49 with NA and NBzA, respectively,
(Figure 12A-c and A-d) at 100 mm concentration and that ob-
tained for control RNA was 0.41 and 0.30 (Figure 12, B-c and B-
d) This tendency was more marked for the translation suppres-
sion of Rluc located upstream of (CAG)89. The activity of Rluc
was 0.75 and 0.71 (Figure 12A-a and A-b, respectively) at the
same ligand concentration, which were significantly higher
than those observed for the translation of Rluc on control RNA
(cf. 0.53 with NA (B-a) and 0.41 with NBzA (B-b)). While the
molecular mechanism was not clear, the presence of (CAG)89
repeat somehow repressed the translation suppression by
both NA and NBzA. The differences between two ligands were
not obvious in the translation of repeat containing RNA.
Figure 11. CD spectra of r(CAG)9 (3 mm) with (a) NA and (b) NBzA (50 mm) in
the presence of 100 mm NaCl (black line), Na2SO4 (red line) or NaClO4 (blue
line) in 10 mm sodium cacodylate buffer (pH 7.0) at room temperature.
2À
creased the intensity in the presence of SO4 anion compared
À
with those in ClÀ and ClO4 anions. The intensity changes were
especially remarkable in the induced CD bands suggesting
that the NA-binding to r(CAG)9 was most favorable in the
medium of increased hydrophobic interaction with 100 mm
Na2SO4 among three conditions. In contrast, the CD spectra of
r(CAG)9 in the presence of NBzA was much less sensitive to
the change from ClÀ to SO42À, and virtually no change from ClÀ
Discussion
In our previous studies on the molecules binding to the G-G
mismatched DNA, the modification of the amide linkage to the
carbamate linkage significantly increased the binding affini-
ty.[29,37] While we anticipated the increased affinity by additional
stacking interaction and flexibility by a longer linker length, it
was not the case for the four molecules NCA, NAC, NCAC, and
NBzAC having a carbamate linkage between two heterocycles.
SPR analysis of NA and five newly synthesized ligands clearly
indicated that only NA and NBzA did show the affinity to
d(CAG)9 and r(CAG)9. Failure of the binding of four ligands with
the carbamate linkage is possibly due to the difference in con-
formations of two heterocycles from those of NA and NBzA,
which altered the required energy for the rearrangement of
two chromophores into the appropriate position for the bind-
À
to ClO4 (Figure 11b). We also looked at the binding of NA
2À
and NBzA by SPR with the buffer containing SO4 anion to
see if there were any effects of counter anion on the binding,
but the SPR signals obtained for r(CAG)9-immobilized sensor
with NA and NBzA up to 3 mm (cf. SPR conditions in Figure 7)
were so weak that we could not discuss the effect of counter
anions.
In vitro translation assay
The effect of NA and NBzA on in vitro translation of genes up-
stream or downstream from (CAG)n repeat RNA was investigat-
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 12. The effect of NA and NBzA on the in vitro translation of RNA template containing Rluc and Fluc genes (A) with or (B) without the intervening se-
quence of (CAG)89 repeat. The activities of reporter proteins were examined in the presence of NA and NBzA, and plotted against ligand concentration. Key:
Rluc activities with (a) NA and (b) NBzA, Fluc activities with (c) NA and (d) NBzA, the ratio of two luciferases (Fluc/Rluc) with (e) NA and (f) NBzA. The number
shown above the bar indicated the ratio of luciferase activity (a–d) and Fluc/Rluc (e and f) relative to those without ligand.
ing. Since NA and NBzA have the same linker structure with
the shortest length among molecules we examined, one of fac-
tors determining the binding to CAG repeat DNA and RNA
would be likely the spatial distance and arrangement of two
chromophores provided by the linker.
to r(CAG)9 without any apparent cooperativity was suggested.
In our previous studies, the large SPR response was rational-
ized by the formation of hairpin secondary structure on the
sensor surface inducing a large change in dielectric constant
near the surface.[27] The large change in SPR response for
d(CAG)9 might be due to the induced formation of hairpin sec-
ondary structure on the surface upon ligand binding. Since
r(CAG)9 is reported to favorably take a hairpin secondary struc-
ture,[19] such ligand-induced hairpin formation may not be con-
ceivable for r(CAG)9.
The SPR sensorgrams obtained for d(CAG)9 and r(CAG)9 with
NA and NBzA were quite different in terms of the magnitude
of the response. The SPR response for d(CAG)9 reaches about
225 RU for the binding of NBzA, whereas that for r(CAG)9 is
only about 13 RU. The amount of immobilized d(CAG)9 and
r(CAG)9 was 497 and 524 RU, respectively, suggesting that the
large difference is not likely due to the amount of immobilized
polynucleotides. Other possibilities are the difference in the
folded structure and the amount of bound ligands.
In addition to the stoichiometry of the binding, the MS anal-
ysis may suggest that NBzA-bound complexes to r(CAG)9 are
thermodynamically more stable than NA-bound complexes if
we could assume a comparable ionization efficiency for both
NA- and NBzA-bound complexes, because the intensity of ions
corresponding to complex containing 2 and 3 NBzA molecules
is higher than those containing NA.
ESI-TOF-MS analysis clearly showed that the binding stoichi-
ometry of NA and NBzA to d(CAG)9 and r(CAG)9 was quite dif-
ferent. A pairwise binding of NA and NBzA was confirmed for
the binding to d(CAG)9, whereas a stepwise binding of ligands
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The results obtained from CD measurements in the different
counter anions showed that the increased hydrophobic inter-
These characteristics of NBzA-binding are due to the presence
of the tricyclic system, by which the stacking interaction with
the neighboring base pairs would likely be promoted. All the
data described here suggested that the linker structure of NA
and NBzA was suitable for the ligand binding to CAG repeat
DNA and RNA based on our molecular design involving two
heterocycles, and the tricyclic system of benzoazaquinolone
does not interfere with binding to the CAG repeat. While the
binding of NBzA was not markedly increased in terms of the
affinity, the tricyclic ring system should provide us with
a larger chemical space than the bicyclic system to design
compounds, which eventually may exhibit exclusive binding to
CAG repeat RNA with high affinity.
2À
action by a salting out SO4 anion promotes the NA-binding
to r(CAG)9 but not the NBzA-binding significantly. This obser-
vation suggests that NBzA intrinsically gained such hydropho-
bic effect but NA-binding was further promoted by the salting
out effect.[39,40] It is reasonable to assume that the expanded
tricyclic system has more chances to gain hydrophobic interac-
tion than the parent two-ring system, the characteristic NBzA-
binding to r(CAG)9 is most likely due to the stacking interaction
of the benzoazaquinolone system with the neighboring bases.
To investigate the effect of NA and NBzA binding to (CAG)n
repeat RNA, we have established the in vitro translation assay,
where two reporter genes, for example, Renilla luciferase (Rluc)
and firefly luciferase (Fluc), were arranged upstream and down-
stream from (CAG)n repeat. Simultaneous suppression of both
Rluc and Fluc expression by ligand suggests that ligand inter-
fered the entire translation system, whereas pronounced sup-
pression of expression of one of two genes could be due to
the presence of (CAG)n repeat.
Experimental Section
Surface plasmon resonance assay for the r(CAG)9-immobilized
surface. A streptavidin-coated sensor chip (SA chip, GE Healthcare)
was washed with HBS-EP buffer (10 mm HEPES pH 7.4, 0.15m NaCl,
3 mm EDTA and 0.005% v/v Surfactant P20) for 6 min and then ac-
tivated with three consecutive 1 min injections of 30 mL activation
buffer (50 mm NaOH and 1m NaCl). 5’-Biotinylated r(CAG)9 (pur-
chased from Gene Design Inc., Osaka Japan) was diluted to 1 mm
with HBS-EP buffer and flowed onto the SA chip at 5 mLminÀ1 for
60 s. The amount of r(CAG)9 immobilized on the chip surface was
524 response units (RU). Binding of the ligands to the surface was
analyzed by using a BIAcore T200 SPR system (GE Healthcare).
For the control RNA that lacked the intervening CAG repeat
sequence between two reporter genes, both NA and NBzA
suppressed the translation of both Rluc and Fluc in concentra-
tion-dependent manner, suggesting that both ligands have in-
hibitory effects on the translation. The ratio of two reporter
genes was 0.78 for NA and 0.73 for NBzA at 100 mm concen-
tration. These ratios are the intrinsic number for the translation
of tandemly arranged Rluc and Fluc under the conditions of
our in vitro translation assay. For the translation of RNA con-
taining CAG intervening repeat sequence between two lucifer-
ases, the effect of ligands on the translation suppression was
less efficient. Especially, the translation of Rluc located up-
stream of CAG repeat was kept in high level of translation
than that in the RNA without CAG repeat (e.g., 0.75 vs. 0.53
for NA, and 0.71 vs. 0.41 for NBzA treatment at 100 mm). The
repressive effect on translation suppression of Rluc gene is due
to the presence of CAG repeat, though the molecular mecha-
nism is not fully understood. Differences in two ligands NA
and NBzA for the translation were not clearly observed. It was
apparent that the tricyclic system in NBzA did not have any
significant positive or negative effects on translation, but pro-
vides large chemical space for further modification to gain
higher affinity and selectivity.
Multi-cycle kinetic analysis was carried out at 258C under the con-
tinuous flow of HBS-EP buffer at a flow rate of 30 mLminÀ1. After
the surface had been conditioned by exposure to the buffer flow
for 60 s, a solution of each concentration of the ligand in HBS-EP
buffer was injected for 60 s to analyze association to the sensor
chip. The buffer was subsequently injected for another 180 s in
order to determine the dissociation of the bound ligand from the
surface.
Single-cycle kinetic analysis was carried out at 258C under the con-
tinuous flow of HBS-EP buffer. To avoid mass transport limitations,
the flow rate of HBS-EP buffer was 90 mLminÀ1 [43]
The surface was
.
conditioned by exposure for 120 s to analyze the association of li-
gands to the sensor chip. Each ligand was dissolved in HBS-EP
buffer at the concentration of 0.75, 1.0, 2.0, 2.5, and 3.0 mm, and
the resulting solutions were sequentially injected over flow cells on
the sensor surface for 120 s at a flow rate of 90 mLminÀ1 in single-
cycle mode. The obtained data were analyzed using the BIAcore
T200 evaluation software, version 2.0, and kinetic parameters were
determined by fitting the data to a 1:1 binding model.
Circular dichromic spectra of r(CAG)9 in the presence of NA and
NBzA. Circular dichroism (CD) measurements were carried out on
a J-725 CD spectropolarimeter (JASCO, Japan) using a 1.0 cm path
length cell. CD spectra of r(CAG)9 (3 mm) in sodium cacodylate
buffer (10 mm, pH 7.0) and NaCl (100 mm) were measured in the
absence and presence of NA or NBzA (50 mm) at 258C.
Conclusions
Among the five new NA derivatives we synthesized, those
modified by the carbamate linkage failed to bind to r(CAG)9.
The NBzA having a tricyclic system of benzoazaquinolone
showed comparable binding affinity and repeat selectivity with
NA, and induced a more marked and intense CD signal change
than NA. ESI-TOF MS analysis revealed that the mode of bind-
ing of NA and NBzA to an r(CAG)n repeat is not the pairwise
binding we observed for the binding to d(CAG)n repeat. The in-
tensity of ion peaks was higher for the NBzA-bound r(CAG)9
than that of NA-bound r(CAG)9, suggesting that NBzA-bound
r(CAG)9 would be more stable than NA-bound complexes.
ESI-TOF MS measurements.[44] ESI-TOF MS experiments were per-
formed on a Bruker maXis impact mass spectrometer in the nega-
tive mode. The source parameters were set as follows: capillary
voltage of +3500 V, nebulizer of 1.0 bar, dry gas flow of 4.0 LminÀ1
at 1608C, and end plate offset voltage of 500 V. The measurements
were operated at a background pressure of 5ꢁ10À7 mbar. The rate
of sample injection into the mass spectrometer was 3 mLminÀ1
.
The mixture of a final concentration of 2 mm r(CAG)9 repeat with
10 mm NA or 10 mm NBzA in 50 mm ammonium acetate solution
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
(pH 7.0) was directly injected into mass spectrometer. Each mass
spectrum was an average of at least 60 scans. Data were analyzed
with the software of Compass DataAnalysis.
thors also thank Platform for Drug Discovery, Informatics, and
Structural Life Science from the Ministry of Education, Culture,
Sports, Science and Technology, Japan for the technical sup-
port on using Molecular modelling software (SCHRꢂDINGER
Suit).
Preparation of RNA template. The pCMVTnT_RFluc vector carrying
Renilla and firefly luciferase reporter gene in tandem downstream
of the T7 promoter sequence was sequentially digested with SalI
and EcoRI (Takara) and purified by using an NucleoSpin Gel and
PCR Clean-up kit (Macherey–Nagel). A dsDNA fragment containing
an EcoNI site was prepared by annealing the two oligonucleotides
5’-AAT TCT CCT GGG GGA GGT G-3’ and 5’-TCG ACA CCT CCC CCA
GGA G-3’ (purchased from Gene Design), producing sticky ends for
SalI and EcoRI. The dsDNA fragment was ligated into the SalI/
EcoRI-digested vector with T4 DNA ligase (New England BioLabs)
to create an EcoNI site between Renilla and firefly luciferase report-
er gene in the pCMVTnT_RFluc vector. The obtained vector was di-
gested with EcoNI (New England BioLabs) and dephosphorylated
with Antarctic Phosphatase (New England BioLabs), followed by
purification with NucleoSpin Gel and PCR Clean-up kit (Macherey–
Nagel). DNA fragments containing the (CAG)n sequence were pre-
pared by ligation of double-stranded DNA containing 5’-G (CAG)9
CA-3’ and 5’-(CTG)10-3’ (purchased from Gene Design) with T4 DNA
ligase (New England BioLabs). The ligation products were separat-
ed by native PAGE followed by staining with SYBRꢀGold, and the
bands ranging from 150 to 300 bp were excised and eluted from
the gel. The obtained DNA fragments were ligated into the EcoNI-
digested vector and the product was then used to transform NEB
Stable Competent E. coli (New England Biolabs). The individual
clones were partially sequenced and the number of CAG repeats
was determined. Among the clones isolated, the clone carrying
(CAG)89 sequence (pCMVTnT_RFluc_CAG89) was used for preparing
the RNA template for in vitro translation. pCMVTnT_RFluc_CAG89
was linearized by digestion with BamHI (Takara) prior to in vitro
transcription. The linearized pCMVTnT_RFluc_CAG89 was tran-
scribed with T7 RNA polymerase (HiScribe T7 Quick High Yield RNA
Synthesis Kit, New England BioLabs) to produce RNA containing
the CAG89 repeat. After in vitro transcription, the reaction mixture
was treated with RNase-free DNase I (QIAGEN) to digest the DNA
template, and precipitated with ammonium acetate/2-propanol.
The precipitates were dissolved in RNase-free water and the mix-
ture was applied to a NAP-5 gel filtration column (GE Healthcare)
to remove unincorporated NTPs. The purified RNA was dissolved in
water, analyzed by 1.5% (w/v) agarose gel electrophoresis, and
stored at À808C.
Keywords: CAG repeats · recognition · RNA · small molecules ·
surface plasmon resonance · trinucleotide repeats
[1] G. R. Wilmot, S. T. Warren in Genetic Instabilities and hereditary neurologi-
cal diseases, (Eds, : R. D. Wells, S. T. Warren), Academic Press, San Diego,
1998, pp. 3–12.
[2] C. J. Cummings, H. Y. Zoghbi, Annu. Rev. Genomics Hum. Genet. 2000, 1,
281–328.
[3] S. M. Mirkin, Nature 2007, 447, 932–940.
[4] H. T. Orr, H. Y. Zoghbi, Annu. Rev. Neurosci. 2007, 30, 575–621.
[5] S. W. Davies, M. Turmaine, B. A. Cozens, M. DiFiglia, A. H. Sharp, C. A.
Ross, E. Scherzinger, E. F. Wanker, L. Mangiarini, G. P. Bates, Cell 1997, 90,
537–548.
[6] F. O. Walker, Lancet 2007, 369, 218–228.
[7] J. D. Brook, M. E. McCurrach, H. G. Harley, A. J. Buckler, D. Church, H.
Aburatani, K. Hunter, V. P. Stanton, J. P. Thirion, T. Hudson, R. Sohn, B,
Zemelman, R. G. Snell, S. A. Rundle, S. Crow, J. Davies, P. Shelbourne, J.
Buxton, C. Jones, V. Juvonen, K. Johnson, P. S. Harper, D. J. Shaw, D. E.
Housman, Cell 1992, 68, 799–808.
[8] H. Jiang, A. Mankodi, M. S. Swanson, R. T. Moxley, C. A. Thornton, Hum.
Mol. Genet. 2004, 13, 3079–3088.
[9] M. B. de Leꢃn, B. Cisneros, J. Neurosci. Res. 2008, 86, 18–26.
[10] Y. H. Fu, D. P. Kuhl, A. Pizzuti, M. Pieretti, J. S. Sutcliffe, S. Richards, A. J.
Verkerk, J. J. Holden, R. G. Fenwick, S. T. Warren, B. A. Oostra, D. L.
Nelson, C. T. Caskey, Cell 1991, 67, 1047–1058.
[11] R. Willemsen, J. Levenga, B. A. Oostra, Clin. Genet. 2011, 80, 214–225.
[12] S. I. Bidichandani, T. Ashizawa, P. I. Patel, Am. J. Human Genet. 1998, 62,
111–121.
[13] M. B. Delatycki, R. Williamson, S. M. Forrest, J. Med. Genet. 2000, 37, 1–
8.
[14] A. R. La Spada, J. P. Taylor, Nat. Rev. Genet. 2010, 11, 247–258.
[15] C. T. McMurray, Nat. Rev. Genet. 2010, 11, 786–799.
[16] V. Dion, J. H. Wilson, Trends Genet. 2009, 25, 288–297.
[17] J. D. Cleary, K. Nichol, Y. H. Wang, C. E. Pearson, Nat. Genet. 2002, 31,
37–46.
[18] K. Ohshima, R. D. Wells, J. Biol. Chem. 1997, 272, 16798–16806.
[19] C. T. McMurray, Proc. Natl. Acad. Sci. USA 1999, 96, 1823–1825.
[20] K. Sobczak, M. de Mezer, G. Michlewski, J. Krol, W. J. Krzyzosiak, Nucleic
Acids Res. 2003, 31, 5469–5482.
[21] C. E. Pearson, R. R. Sinden, Biochemistry 1996, 35, 5041–5053.
[22] S. Kang, A. Jaworski, K. Ohshima, R. D. Wells, Nat. Genet. 1995, 10, 213–
218.
[23] M. J. Hartenstine, M. F. Goodman, J. Petruska, J. Biol. Chem. 2000, 275,
18382–18390.
In vitro translation assay. In vitro translation was performed using
Rabbit Reticulocyte Lysate System (Promega). Translation reactions
(10 mL) contained 70% lysate, 20 mm amino acid mixture, and
85 fmol template RNA in the presence of the ligand at the final
concentration of 0, 1, 5, 10, 25, 50, and 100 mm. The reaction mix-
tures were incubated at 308C for 60 min and were quenched by
´
[24] A. Marquis Gacy, G. Goellner, N. Juranic, S. Macura, C. T. McMurray, Cell
1995, 81, 533–540.
adding RNase A (10 mL) to a final concentration of 0.2 mgmLÀ1
.
[25] C. E. Pearson, R. R. Sinden, Curr. Opin. Struct. Biol. 1998, 8, 321–330.
[26] C. E. Pearson, M. Tam, Y. H. Wang, S. E. Montgomery, A. C. Dar, J. D.
Cleary, K. Nichol, Nucleic Acids Res. 2002, 30, 4534–4547.
[27] K. Nakatani, S. Hagihara, Y. Goto, A. Kobori, M. Hagihara, G. Hayashi, M.
Kyo, M. Nomura, M. Mishima, C. Kojima, Nat. Chem. Biol. 2005, 1, 39–
43.
Firefly and Renilla luciferase activities were measured with a Mithras
LB940 multimode microplate reader using the Dual-Glo Luciferase
Assay System (Promega).
[28] M. Hagihara, H. He, K. Nakatani, ChemBioChem 2011, 12, 1686–1689.
[29] T. Peng, K. Nakatani, Angew. Chem. Int. Ed. 2005, 44, 7280–7283; Angew.
Chem. 2005, 117, 7446–7449.
Acknowledgements
[30] M. Hagihara, H. He, M. Kimura, K. Nakatani, Bioorg. Med. Chem. Lett.
2012, 22, 2000–2003.
This work was supported by JSPS KAKENHI Grant-in-Aid for
Specially Promoted Research (26000007) and partially by the
Advanced Research for Medical Products Mining Program of
the National Institute of Biomedical Innovation of the National
Institute of Biomedical Innovation (NIBIO, 10–22) for KN. Au-
[31] T. Zu, B. Gibbens, N. S. Doty, M. Gomes-Pereira, A. Huguet, M. D. Stone,
J. Margolis, M. Peterson, T. W. Markowski, M. A. Ingram, Z. Nan, C. For-
ster, W. C. Low, B. Schoser, N. V. Somia, H. B. Clark, S. Schmechel, P. B.
Bitterman, G. Gourdon, M. S. Swanson, M. Moseley, L. W. Ranum, Proc.
Natl. Acad. Sci. USA 2011, 108, 260–265.
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[32] C. E. Pearson, Plos Genet. 2011, 7, e1002018.
[33] M. Wojciechowska, M. Olejniczak, P. Galka-Marciniak, M. Jazurek, W. J.
Krzyzosiak, Nucleic Acids Res. 2014, 42, 11849–11864.
[40] R. Breslow, C. J. Rizzo, J. Am. Chem. Soc. 1991, 113, 4340–4341.
[41] B. A. Sherf, S. L. Navarro, R. R. Hannah, K. V. Wood, Promega Notes 1996,
57, 2–9.
[34] K. Nakatani, S. Sando, I. Saito, Nat. Biotechnol. 2001, 19, 51–55.
[35] K. Nakatani, S. Sando, H. Kumasawa, J. Kikuchi, I. Saito, J. Am. Chem.
Soc. 2001, 123, 12650–12657.
[36] K. Nakatani, Bull. Chem. Soc. Jpn. 2009, 82, 1055–1069.
[37] T. Peng, T. Murase, Y. Goto, A. Kobori, K. Nakatani, Bioorg. Med. Chem.
Lett. 2005, 15, 259–262.
[38] S. Hagihara, H. Kumasawa, Y. Goto, G. Hayashi, A. Kobori, I. Saito, K. Na-
katani, Nucleic Acids Res. 2004, 32, 278–286.
[39] W. D. Ding, G. A. Ellestad, J. Am. Chem. Soc. 1991, 113, 6617–6620.
[42] A. Siddiqui-Jain, C. L. Grand, D. J. Bearss, L. H. Hurley, Proc. Natl. Acad.
Sci. USA 2002, 99, 11593–11598.
[43] R. Karlsson, A. Fꢄlt, J. Immunol. Methods 1997, 200, 121–133.
[44] L.-P. Bai, M. Hagihara, K. Nakatani, Z.-H. Jiang, Sci. Rep. 2014, 4, 6767.
Manuscript received: April 16, 2016
Accepted Article published: May 16, 2016
Final Article published: && &&, 0000
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
Binding to Trinucleotide Repeats
Repeat RNA Binding: Structure-binding
studies of a series of naphthyridine-aza-
quinolone derivatives are described.
NBzA with the tricyclic system of ben-
zoazaquinolone induced a more exten-
sive conformational change on the CAG
repeat. The linker structure in NA and
NBzA was suitable for the binding to
CAG DNA and RNA, and the tricyclic
benzoazaquinolone did not interfere
with the binding.
Jinxing Li , Akihiro Sakata, Hanping He,
Li-Ping Bai, Asako Murata,
Chikara Dohno, Kazuhiko Nakatani*
&& – &&
Naphthyridine-Benzoazaquinolone:
Evaluation of a Tricyclic System for
the Binding to (CAG)n Repeat DNA and
RNA
&
&
Chem. Asian J. 2016, 00, 0 – 0
www.chemasianj.org
12
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!