Ametantrone-based compounds as potential regulators of Tau pre-mRNA alternative splicing

Article abstract of DOI:10.1039/c4ob01925c

Tau pre-mRNA contains a stem-loop structure involved in the regulation of the alternative splicing of tau protein. We describe here a new family of Tau RNA ligands selected by dynamic combinatorial chemistry based on the combination of ametantrone with small RNA-binding molecules. The most promising compound results from derivatization of one of the side chains of the anthraquinone ring with the small aminoglycoside neamine through a short spacer. This compound binds the RNA target with a high affinity in a preferred binding site, in which the heteroaromatic moiety intercalates in the bulged region of the stem-loop and its side chains and neamine interact with the major groove of the RNA. Importantly, binding of this compound to mutated RNA sequences involved in the onset of some tauopathies such as FTDP-17 restores their thermodynamic stability to a similar or even higher levels than that of the wild-type sequence, thereby revealing its potential as a modulator of Tau pre-mRNA splicing.

Full text of DOI:10.1039/c4ob01925c

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Ametantrone-based compounds as potential
regulators of Tau pre-mRNA alternative splicing†
Cite this: DOI: 10.1039/c4ob01925c
a
a
b
a
Gerard Artigas, Paula López-Senín, Carlos González, Núria Escaja and
a
Vicente Marchán*
Tau pre-mRNA contains a stem–loop structure involved in the regulation of the alternative splicing of tau
protein. We describe here a new family of Tau RNA ligands selected by dynamic combinatorial chemistry
based on the combination of ametantrone with small RNA-binding molecules. The most promising com-
pound results from derivatization of one of the side chains of the anthraquinone ring with the small
aminoglycoside neamine through a short spacer. This compound binds the RNA target with a high affinity
in a preferred binding site, in which the heteroaromatic moiety intercalates in the bulged region of the
stem–loop and its side chains and neamine interact with the major groove of the RNA. Importantly,
binding of this compound to mutated RNA sequences involved in the onset of some tauopathies such as
FTDP-17 restores their thermodynamic stability to a similar or even higher levels than that of the wild-type
sequence, thereby revealing its potential as a modulator of Tau pre-mRNA splicing.
Received 10th September 2014,
Accepted 22nd October 2014
DOI: 10.1039/c4ob01925c
www.rsc.org/obc
patients suffering from neurodegenerative diseases, including
Alzheimer’s disease and some tauopathies. Alternative spli-
Introduction
5
RNA molecules, not only mRNAs or pre-mRNAs but also the cing of exons 2, 3, and 10 of the MAPT gene, which encodes
wide variety of non-coding RNAs (e.g. miRNAs or their pre- for tau protein, gives rise to six isoforms with either three (3R)
miRNA precursors), are therapeutically relevant targets in med- or four (4R) repeat domains, which differ in their relative
1
icinal chemistry. Like proteins, RNA folds local secondary ability to bind to microtubules. In general, the relative ratio of
structures (e.g. stem–loops, bulges, internal loops, etc.) into 4R to 3R is close to 1 in healthy adult human brains. However,
complex three-dimensional architectures, generating unique this ratio is found altered in many tauopathies. The impor-
2
binding sites where small molecules can be accommodated.
tance of a well-defined stem–loop structure located at the exon
Owing to the ubiquity of RNA-mediated biological processes, 10-5′ intron junction of Tau pre-mRNA as a regulatory element
targeting therapeutically relevant RNA sequences with small of alternative splicing has been established both in vitro and
6
molecules offers new opportunities to control RNA function in vivo. Such studies revealed that the extent of exon
3
and, for instance, to treat human diseases. Despite the pro- 10 inclusion is inversely proportional to the stability of this
gress made in recent years, the discovery of ligands with high local secondary structure. Some intronic mutations (named +3,
affinity and sequence selectivity for a given RNA target and +13, +14 and +16 in Scheme 1) found in patients with tauo-
with desirable drug-like properties still remains a challenge. pathies, such as frontotemporal dementia with parkinsonism
This is mainly a consequence of our poor understanding of linked to chromosome 17 (FTDP-17), decrease the stem–loop
7
RNA–ligand recognition principles, which hinders the rational thermodynamic stability, leading to an increase in exon
3
b,4
design of new compounds based on structural studies.
The microtubule-associated protein tau is required for the pared to the 3R isoforms.
polymerization and stability of microtubules, as well as for From a therapeutic point of view, it has been hypothesized
10 inclusion and an overproduction of 4R tau isoforms com-
6
,7
maintaining neuronal integrity and axonal transport. Neuronal that ligands that target the RNA secondary structure located
filamentous deposits of this protein have been found in at the exon 10-5′ intron junction of Tau pre-mRNA would
have the capacity to restore the physiological balance of
tau isoforms generated upon abnormal alternative splicing
a
Departament de Química Orgànica and IBUB, Universitat de Barcelona, Martí i
Franquès 1-11, E-08028 Barcelona, Spain. E-mail: vmarchan@ub.edu;
Fax: (+) (34) 93 339 7878
by increasing the thermodynamic stability of the mutated
sequences. Moreover, such compounds would allow one to
study how local secondary RNA structures can influence
b
Instituto de Química Física Rocasolano, CSIC, Serrano 119, E-28006 Madrid, Spain
8
the outcome of pre-mRNA splicing. In recent years, several
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4ob01925c
ligands able to bind and stabilize wild type and mutated Tau
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 1 (A) Sequences and secondary structure of wild-type (wt) and +3, +13, +14 and +16 mutated Tau stem–loop RNAs. Exonic sequences are
shown in capital letters and intronic sequences in lower case. Nucleotides involved in base pairs, previously identified by NMR, are connected by a
6
a,11
dash.
When required, biotin or fluorescein derivatization was performed at the 5’ end. The ends of the chains were modified with 2’-O-methyl-
13a
ribonucleosides (denoted by an asterisk). (B) Structure of mitoxantrone, ametantrone and acridine–neamine ligands. (C) Schematic representation
of the derivatization of ametantrone through one of the side chains of the anthraquinone ring.
1
1
stem–loop structures have been reported. Varani and collabor- According to the reported structural data, the phenolic
ators described the three-dimensional structure of the RNA hydroxyl groups of mitoxantrone do not seem to have any
9
11,12
complex with the aminoglycoside neomycin B. More recently, important role in the recognition of Tau RNA.
Conse-
the anticancer drug mitoxantrone (Mtx in Scheme 1) was quently, instead of mitoxantrone, we selected ametantrone, an
identified by Wolfe and collaborators in a high-throughput anthraquinone analogue (Amt in Scheme 1) with a proven
1
0
15
screening as a promising ligand. Subsequent NMR structural lower cytotoxicity in eukaryotic cells. Since both side chains
studies of the complex of mitoxantrone with Tau RNA revealed of mitoxantrone play an important role in the recognition of
that the anthraquinone ring of this drug intercalates between the stem–loop structure, in particular the ammonium func-
1
1
11
the two G : C base pairs flanking the bulged adenine. In tions, we planned to use the end of one of the side chains of
addition, the evaluation of the binding to Tau RNA and the ametantrone as a derivatization point to attach the amino-
stabilization capacity of a series of mitoxantrone analogues glycoside moiety. Hence, we report here new Tau RNA ligands
allowed the establishment of the structure–activity based on the derivatization of ametantrone with small
1
2
relationships.
In previous studies,
RNA binding molecules (azaquinolone and neamine)
we used dynamic combinatorial (Scheme 1C), together with several spectroscopic studies
1
3
1
4
chemistry (DCC) to identify a series of ligands that bind the (UV-Vis, fluorescence and NMR) on their interaction with the
Tau stem–loop structure with high to medium affinities, RNA target.
showing the desired ability to stabilize both wt and mutant
RNA sequences upon binding. The most promising RNA-
templated ligands combined planar heteroaromatic rings such
as acridine with neamine (Acr–Nea/Nea2 in Scheme 1B),
a small aminoglycoside that incorporates rings I and II of
Results and discussion
Selection of Tau RNA ligands
1
3a
neomycin, which are important structural motifs in RNA In general, there is good correlation between the ligand’s
recognition. Guanidinylation of the neamine moiety had an amplification in dynamic combinatorial chemistry experi-
important effect, improving not only the binding affinity of the ments and its binding affinity for a given biological target,
1
4
ligands, but also their ability to stabilize the mutated sequen- either a nucleic acid (DNA or RNA) or a protein. Hence, prior
1
3b
ces.
Biophysical studies showed that these ligands interact to synthesizing ametantrone–neamine ligands, we planned to
with RNA via intercalation of the acridine moiety, probably carry out a DCC experiment to validate our hypothesis using
at the bulged region of the stem–loop structure, and that thiol-derivatized building blocks. The ligand’s amplification
the combination with the aminoglycoside improves the RNA might provide, in addition, valuable information about the
1
3
binding properties of the heteroaromatic moiety.
optimal length of the spacer linking the two moieties. For this
On the basis of all these precedents, we wondered whether purpose, compounds with different distances between the
replacement of acridine by mitoxantrone in acridine–neamine reactive functional group and the building block core, such as
compounds would lead to Tau pre-mRNA ligands with the two neamine monomers shown in Fig. 1A, have been
enhanced RNA-binding properties and, more importantly, tested.
whether ligands incorporating this anthraquinone moiety
would maintain the preferred binding site of mitoxantrone at exchange reactions,
the bulged region of the stem–loop structure of Tau RNA. monomer, Amt–SH (Scheme 2), was required. We decided to
To perform DCC experiments using thiol–disulfide
a
suitably-modified ametantrone
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 1 Results of a DCC experiment involving 5’-biotinylated wt Tau RNA and the Amt, Acr, Nea, Nea2, Azq and TyrP thiol-derivatized monomers.
(
(
A) Structure and peptide sequence of the building blocks. (B) HPLC traces showing the composition of the DCL in the absence (left) and presence
right) of RNA after 48 h. (C) Histograms showing the changes in the DCL composition (left) and the percentage changes (% amplification) of each
species (right) in the presence of wt Tau RNA. (D) Structure of the selected ligands in the DCC experiment.
replace the hydroxyl group of one of the side chains of ametan- thioacetate derivative (2) as a major product. Finally, Amt–SH
trone by a thiol group since this small modification was was obtained after hydrolysis of the thioester group with
expected not to disturb the interaction with RNA. Ametantrone sodium methoxide followed by an acidic treatment with TFA
(Amt) was first synthesized following a reported procedure in the presence of TIS and water as scavengers to remove the
by reacting 1,4-difluoroanthraquinone with an excess of Boc protecting groups.
1
6
N-(2-hydroxyethyl)ethylenediamine. After selective Boc-pro-
The DCC experiment was carried out as previously repor-
1
3a
tection of the secondary alkyl amino functions, a Mitsunobu ted.
Briefly, the thiol-derivatized monomers (Amt–SH,
3
reaction of 1 with DIAD, PPh and thiolacetic acid afforded the Acr–SH, Nea–SH, Nea2–SH, Azq–SH and TyrP–SH; see Fig. 1A
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 2 Synthesis of the thiol-derivatized ametantrone monomer (Amt–SH).
for their structures) were incubated with the biotinylated wt reaction of the intermediates with the ametantrone–thiol
Tau RNA in aqueous Tris–HCl buffer (50 mM, pH 7.7) contain- monomer (Amt–SH), the expected disulfide heterodimers were
ing NaCl (100 mM) and EDTA (0.1 mM) at room temperature isolated by reversed-phase HPLC (yields 33–55%) and fully
under an air atmosphere without stirring. These conditions characterized by MS and NMR.
ensure the correct folding of the RNA target as well as the
In order to validate the results from the DCC experiment,
thiol–disulfide exchange (21 theoretical compounds may be the binding affinities of the selected ligands (Amt–Nea/Nea2
formed). After the isolation of RNA and its interacting ligands and Amt–Azq) for Tau RNA were determined by fluorescence
with streptavidin anchored to magnetic beads, a hot acidic titration experiments using the 5′-end fluorescein-labelled
1
4e
aqueous solution was used to denature it and to provoke the oligoribonucleotide.
As shown in Fig. 2A, the fluorescence
release of RNA-bound ligands. All the compounds were identi- of RNA was quenched on the addition of increasing concen-
fied and quantified by UV-MS-HPLC (see Fig. 1B). The com- trations of the ligand. A characteristic dose-dependent curve
parison with the control experiment (RNA free) allowed the was obtained when the normalized fluorescence intensity at
determination of the % of amplification of the compounds 517 nm was plotted against the compound concentrations. In
in the presence of Tau RNA. As shown in Fig. 1C, disulfide all cases, the full titration was repeated in the absence of RNA
heterodimers incorporating the ametantrone fragment were to subtract the inherent fluorescence of the ligands from that
amplified in a much higher ratio than those containing the of the labelled RNA. EC50 values, which can be assimilated
acridine fragment, particularly when this heteroaromatic com- to dissociation constants, were obtained by fitting the data to a
pound was combined with the neamine monomer: Amt–Nea sigmoidal dose–response curve (Fig. 2B).
(
≈1600%) and Amt–Nea2 (≈3500%) vs. Acr–Nea (≈50%) and
The EC50 values of the ametantrone-containing ligands
Acr–Nea2 (≈150%) (see Fig. 1D and Scheme 1, respectively, for together with those of ametantrone, mitoxantrone and
their structures). This clearly indicated that ametantrone-con- neamine are shown in Table 1. In good agreement with DCC
taining ligands should have a much higher binding affinity results, the most amplified compounds exhibited the highest
1
3a
than those containing acridine. Interestingly, the amplifica- binding affinities. Indeed, ligands containing the ametantrone
tion of Amt–Nea2 was about 2.2-fold higher than that of Amt– fragment showed EC50 values that fall into the nanomolar
Nea. This result reproduced the tendency previously found for scale (EC50 = 78.8 and 70.6 nM for Amt–Nea and Amt–Nea2,
Acr–Nea/Nea2 ligands, indicating that the shortest distance respectively). It is also noticeable that ametantrone–neamine
between the heteroaromatic fragment (acridine or ametan- ligands bind Tau RNA much more strongly than any of their
trone) and the small aminoglycoside neamine is preferred in components alone (see the neamine and ametantrone entries
this family of Tau RNA ligands.
in Table 1), which indicates that the derivatization of one of
On the basis of DCC results, we planned the synthesis of the ametantrone side chains with a polycationic aminoglyco-
the Amt–Nea and Amt–Nea2 ligands on a higher scale to study side has a positive effect on binding affinity. Moreover, the
their interaction with Tau RNA. Although Amt–Azq (Fig. 1D) binding was observed to be approximately two times stronger
was amplified to a much lesser extent (≈200%) than ametan- for Amt–Nea/Nea2 ligands than for mitoxantrone or Amt–Azq.
trone–neamine compounds, it was also selected because of the Interestingly, a correlation was observed between the binding
potential of the azaquinolone fragment to recognize the affinity and the linker length: the binding of Amt–Nea2
1
3a
bulged adenine in the stem–loop.
The synthesis of Amt– was slightly higher than that of the analogous compound
Nea/Nea2 and Amt–Azq was carried out in solution in a two- with the longer spacer linking the ametantrone and
step process. First, one of the thiol-containing monomers neamine moieties (Amt–Nea). This is consistent with % ampli-
(
(
Nea/Nea2–SH or Azq–SH) was activated with 2,2′-dithiobis- fication in DCC and with previous studies on Acr–Nea/Nea2
5-nitropyridine) (DTNP) in a slightly acidic medium. After ligands.
1
7
13a
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 (A) Fluorescence quenching of wt Tau RNA labelled with fluorescein on the addition of increasing concentrations of Amt–Nea2. Experi-
mental conditions: [RNA] = 25 nM and [ligand] = 0 to 614 nM, in 10 mM sodium phosphate buffer pH 6.8, 100 mM NaCl and 0.1 mM Na EDTA. (B)
Plot of the normalized fluorescence signal at 517 nm against the log of the ligand concentration. The normalized fluorescence was calculated by
dividing the difference between the observed fluorescence, F’, and the final fluorescence, F , by the difference between the initial fluorescence, F
and the final fluorescence, F . Hence, absolute values of the fluorescence changes upon ligand titration are shown.
2
f
0
,
f
1
8b
Table 1 Binding of the ligands to wt RNA in the absence or in the pres- previously reported by Tor et al.
with ligands for the HIV-1
ence of a tRNA competitor
RRE RNA target based on the combination of acridine with
neomycin, but differs from that we found with Acr–neamine
EC50 b(nM) +
EC50 + tRNA /
EC50
1
3a
a
ligands.
The overall results indicated that ametantrone not
Ligand
EC50 (nM)
tRNA
only generates ligands with higher affinity for Tau RNA than
acridine upon combination with neamine, but that this
anthraquinone derivative confers higher specificity on them,
particularly when comparing ligands with the shortest spacer.
6
Neamine
3.1 × 10
nd
nd
4.8
2.9
3.1
9.1
8.1
Mitoxantrone
Ametantrone
Amt–Azq
Amt–Nea
Amt–Nea2
168.8 ± 6.2
231.8 ± 8.0
162.5 ± 5.7
78.8 ± 6.1
70.6 ± 7.2
803.3
675.6
570.7
714.2
569.8
Spectroscopic studies on the interaction of Amt-containing
ligands with Tau RNA
a
Fluorescence measurements were performed in 10 mM sodium
EDTA, with
RNA] = 25 nM. Measured in the presence of a 30-fold nucleotide
phosphate buffer pH 6.8, 100 mM NaCl and 0.1 mM Na
[
excess of a mixture of tRNA (tRNA ).
2
b
mix
Our next objective was focused on the study of the interaction
of ametantrone-containing ligands with Tau RNA using several
spectroscopic techniques (UV-Vis, fluorescence and NMR) in
One of the main problems of RNA ligands based on small order to assess whether this new family of compounds has a
molecules is the achievement of sequence selectivity, some- preferred binding site in the target. As previously stated for
1
8
times referred to as specificity. This is because of the rela- mitoxantrone, the presence of a bulged region in Tau RNA,
tively low cellular expression level of disease-associated RNAs where adenine is flanked by two G : C pairs, facilitates the
in comparison with ribosomal RNAs, which represent about intercalation of the anthraquinone ring, causing the displace-
1
9
8
0–90% of the total RNA content of living cells. Hence, the ment of the bulged adenine and the establishment of
ratio between the EC50 values in the presence and in the sequence-specific contacts between the positively-charged side
mix
11
absence of a large excess of tRNA
from baker’s yeast might chains and the RNA nucleobases and phosphate groups.
provide an idea of the specificity of the ligand. As shown in
First, the UV-visible absorption spectra and the fluore-
Table 1, ametantrone was found to be slightly more specific scence emission spectra (upon excitation at 547 nm) were
for Tau RNA than mitoxantrone. This result together with its recorded for all the compounds (Amt, Amt–Azq and Amt–Nea/
1
5
reported low cytotoxicity supports the initial choice of ame- Nea2, see Fig. 3 and Fig. S2 and S3 in the ESI†) in the absence
tantrone instead of mitoxantrone for developing new Tau RNA and in the presence of increasing amounts of wt Tau RNA. As
ligands. As expected, the combination of ametantrone with expected, strong hypochromic and bathochromic effects were
neamine afforded less specific ligands than the combination observed in the UV-Vis spectra during the titration, which are
with azaquinolone. In the presence of the competitor, the EC50 consistent with the binding of the heteroaromatic moiety
values of Amt–Nea and Amt–Nea2 for wt RNA were increased through an intercalative mechanism. The intensity of the two
by 9-fold and 8-fold, respectively, whereas that of Amt–Azq was bands in the 450–700 nm region decreased gradually until sat-
only increased by 3-fold. Interestingly, the specificity of uration was reached (about 3.9 mol equiv. of RNA), with two
the ametantrone–neamine ligands was significantly higher maxima at 594 and 642 nm (overall redshift of 12 and 16 nm,
than that previously found for acridine–neamine ligands, respectively). A similar behaviour was found with Amt–Azq and
particularly in the case of the compound containing the Amt–Nea/Nea2, although the hypochromic and bathochromic
shorter linker. Comparison of specificity ratios for Amt–Nea effects were slightly more pronounced. Fluorescence titration
and Amt–Nea2 indicates that a shorter spacer provides experiments reproduced these results, although the quenching
a slightly higher specificity. This trend is similar to that of the fluorescence was slightly higher for the ligands with
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 3 UV-Vis spectra (A) and fluorescence emission spectra (B) of Amt–Nea2 in the absence and in the presence of increasing amounts of wt RNA
in 10 mM sodium phosphate buffer, pH 6.8, containing 100 mM NaCl and 0.1 mM Na EDTA. 8.4 µM and 2 µM solutions of Amt–Nea2 were used in
2
UV-Vis and fluorescence titration experiments, respectively. The emission spectra were recorded from 600–850 nm with λex = 547 nm.
smaller EC50 values, showing a parallelism with the binding
affinity.
Although these results are consistent with the intercalation
of the anthraquinone chromophore in Tau RNA, the compari-
1
1
son with the behaviour previously reported for mitoxantrone
and some acridine-containing ligands
1
3a
seems to suggest
some differences in their binding mode. This might be attribu-
ted to a higher affinity for a specific and well-defined binding
site in the Tau RNA structure. As a consequence, these ligands
would be allocated at their preferred binding site in the
complex at a lower RNA/ligand ratio. More detailed infor-
mation on the mode of interaction of ametantrone-containing
2
0
ligands was obtained by NMR spectroscopy.
Titration experiments of the ligands (Amt, Amt–Azq and
Amt–Nea/Nea2) with Tau RNA were monitored by NMR (see
Fig. 4A and Fig. S4 and S5 in the ESI†). Addition of Amt or
Amt–Azq caused a general line broadening of all RNA signals,
in particular those close to the bulged adenine (U0 and G−1).
In contrast, titration with Amt–Nea or Amt–Nea2 provoked very
little line broadening, even in excess of ligand (Fig. 4 and S5†).
This different behaviour, and the observation of only one set
of signals under 1 : 1 complex conditions, suggests that Amt–
Nea/Nea2 binds to the target in a more defined binding site
than the other compounds. As can be seen in Fig. 4A, at a
RNA : Amt–Nea2 ratio of 1 : 0.5, different imino signals arising
from the free and bound RNA could be observed. This indi-
cates that both species are in slow equilibrium on the NMR
time scale. 2D NMR experiments were carried out with the 1 : 1
complex Amt–Nea2 : RNA. Provided that the RNA concentration
does not exceed 300 µM to avoid aggregation of the oligoribo-
nucleotide, the resulting 2D spectra of Tau RNA and of the
complex with the ligand were of sufficient quality to proceed
with a more in-depth structural analysis. Some preliminary
Fig. 4 (A) Imino region of the NMR spectra of wt Tau RNA alone and in
the presence of increasing amounts of Amt–Nea2. From the bottom to
the top: ligand/RNA ratio = 0.0, 0.5, 1.0, 1.5 and 2.0. Assignments in gray
1
assignments could be obtained with standard H experiments
11
(NOESY, TOCSY and DQF-COSY), facilitated by previous assign- are labelled according to Varani et al. Experimental conditions: [RNA] =
1
00 μM, 10 mM phosphate buffer in H
5 °C. (B) Exchangeable proton region of the NOESY spectrum (tm
00 ms) of the 1 : 1 complex Amt–Nea2 : wt Tau RNA in H O–D O 9 : 1
2 2
O–D O 90 : 10, pH = 6.8, T =
ments on Tau RNA and the Mtx–Tau RNA complex reported by
6
a,11
=
Varani’s group.
To identify the region of the stem–loop
1
2
2
structure more affected by the ligand, the exchangeable proton
region was first examined since it exhibits a favourable chemi-
in 10 mM phosphate buffer, pH = 6.8, T = 5 °C. [Tau RNA] = 280 μM.
RNA and Amt–Nea2 proton signals are labelled according to the num-
cal shift dispersion (Fig. 4 and S6†). Upon complex formation, bering schemes shown in Scheme 1 and Fig. 1D, respectively.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
the most pronounced changes in chemical shifts involved
imino protons of G−1 and U0, which were shifted by −0.88
and 0.36 ppm, respectively, compared with their values in the
free RNA. In contrast, imino signals of G+1 (Δδ = +0.1 ppm),
U+2 (Δδ = +0.05 ppm) and G+3 (Δδ = −0.03 ppm) were not sig-
nificantly affected in the complex. The imino proton signals of
U+11 and G+5 were rather broad in the free RNA and were not
clearly detected in the complex. Interestingly, a new imino
signal at 12.50 ppm appeared in the complex, which was
assigned to G+17. This suggests that the C−3 : G+17 base pair
is probably more dynamic and exposed to the solvent in the
isolated RNA than in the complex with Amt–Nea2.
Regarding the ligand, some diagnostic signals were also
easily distinguished in the NMR spectra of the complex. As
shown in Fig. 4, two signals in the exchangeable proton region
(
δ = 9.92 and 9.52 ppm) corresponding to the amino protons
NH(11) and NH(18) (see Fig. 1D for the numbering) of the two
side chains of the ametantrone moiety of the ligand were
clearly observed. The two amino signals show numerous cross-
peaks with other protons of the ligand (aromatic protons H2
Fig. 5 Model of the 1 : 1 complex Tau RNA : Amt–Nea2. Coordinates for
11
and H3 and methylene protons of the side chains) and, impor- the Tau RNA : Mtx complex determined by Varani et al. were used to
build the model, as described in the Experimental section.
tantly, with some exchangeable and non-exchangeable
protons of RNA. A number of key intermolecular cross-peaks
could be assigned: NH(18)Amt–H2Amt/H3Amt/H41C+16/
neamine moiety lie in the major groove of the double helical
H42C+16/H8G+17/H1G+17,
NH(11)-H2Amt/H3Amt/H8G−1
segment. The extensive contacts between the side chains of
Amt and the spacer with the RNA probably contribute to the
enhanced binding affinity of this ligand for Tau RNA and
favour the formation of a unique complex in contrast to that
and H41C+16-H2Amt/H3Amt (Fig. 4B and S6 in the ESI†). In
addition, numerous cross-peaks between the Amt–Nea2 side
chain protons and the aromatic protons of the RNA were
observed (Fig. S7†). Although at this stage the sequential
assignment of the later ones is still ambiguous, these contacts
indicate that the side chains interact with the RNA through
the major groove. Most neamine protons are in the
11
observed for the complex with mitoxantrone.
As an additional check of the preference of ametantrone–
neamine ligands for the Tau RNA bulged region, we explored
the interaction of this ligand with two variants of the A-site
RNA fragment: the native sequence, which contains an
internal bulged region where aminoglycosides can be accom-
modated, and a mutant sequence lacking the internal bulged
adenine. As shown in Fig. S13 (see the ESI†), the imino region
of the NMR spectra of either the A-site sequence or its mutant
analogue did not change significantly in the presence of
increasing amounts of Amt–Nea ligands (up to 1.5 mol equiv.).
Moreover, no signals from the ametantrone side chain amino
groups were observed in the exchangeable proton region of
these spectra. These results clearly differ from those obtained
with Tau RNA (see Fig. 3), supporting our conclusion that the
bulged region of Tau RNA is a preferred binding site for the
Amt–Nea ligands described in this study.
3
.3–4.1 ppm region, but no intermolecular NOEs with the RNA
could be found. However, the sign and intensity of NOE cross-
peaks between neamine protons indicate that this moiety
maintains the same correlation time as the rest of the
complex, and is not dangling independently. Most probably,
the neamine interacts mainly with the ribose–phosphate back-
bone of the RNA, giving rise to intermolecular cross-peaks that
lie in very crowded regions of the NOESY spectra.
In summary, NMR data are consistent with a mode of inter-
action in which the ametantrone moiety of the ligand interacts
with the bulged region of the RNA stem–loop structure, most
1
1
probably occupying the same site as mitoxantrone. The struc-
tural determination of the complex Tau RNA : Amt–Nea2 is
outside the scope of this work, since it requires a complete reso-
nance assignment and the extraction of a full set of distance
1
3
15
Effect of Amt-containing ligands on the thermal stability of
Tau RNA mutated sequences
constraints, only possible with isotopic-labelling ( C, N and
probably partial deuteration of riboses). Such a study is now in
progress and will be presented in due course. However, a feas- After establishing that the derivatization of ametantrone with
ible model of the complex can be spotted on the basis of the neamine generates ligands for Tau RNA with a high binding
previously reported complex between Tau RNA and mitoxan- affinity and a clear preference for the bulged region of
11
trone
and the experimental information obtained here the stem–loop structure, we focused on the evaluation of
(
Fig. 5 and S8†). In this model, the anthraquinone ring is the impact of these compounds on the thermal stability of Tau
1
4e
intercalated between G(−1) : C(+16) and C(−3) : G(+17) base RNA by UV-monitored melting experiments.
As previously
pairs, and the two side chains of ametantrone and the stated, effective ligands for Tau RNA are expected to reverse
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 2 Melting temperatures (T
m
, °C) for the complexation of the ligands with Tau RNAs (1 µM both in RNA and in ligands in 10 mM sodium phos-
2
phate buffer, pH 6.8, 100 mM NaCl and 0.1 mM Na EDTA)
a
m
a
m
a
a
T
m
wt
ΔT
T
m
+ 3
ΔT
T
m
+ 14
ΔT
m
T
m
+ 16
ΔT
m
No ligand
Neamine
Mitoxantrone
Ametantrone
Amt–Azq
Amt–Nea
Amt–Nea2
66.5 ± 0.1
66.9 ± 0.3
69.2 ± 0.4
67.1 ± 0.3
67.8 ± 0.6
72.9 ± 0.2
75.4 ± 0.4
—
51.4 ± 0.3
51.5 ± 0.5
57.1 ± 0.5
54.2 ± 0.5
56.1 ± 0.6
60.7 ± 0.4
64.3 ± 0.5
—
53.5 ± 0.3
54.0 ± 0.4
58.1 ± 0.4
55.7 ± 0.6
57.3 ± 0.3
60.7 ± 0.6
63.9 ± 0.5
—
59.6 ± 0.2
60.1 ± 0.3
63.4 ± 0.6
61.6 ± 0.4
61.8 ± 0.4
66.5 ± 0.5
69.4 ± 0.4
+0.4
+2.7
+0.5
+1.3
+6.4
+8.9
+0.1
+5.7
+2.8
+4.7
+9.3
+12.9
+0.5
+4.6
+2.2
+3.8
+7.2
+10.4
+0.5
+3.8
+2.0
+2.2
+6.9
+9.8
a
m m m
ΔT = (T of the RNA in the presence of ligand) − (T of RNA alone).
the destabilization of the stem–loop structure caused by
mutations associated with FTDP-17. The +3, +14 and
+
16 mutated sequences (see Scheme 1) were considered in this
study since they cause significant destabilization with respect
to the wt sequence (−15 °C, −13 °C and −7 °C, respectively).
As shown in Table 2 and Fig. S9–S12 (see the ESI†), the ΔT
m
values for the RNA sequences in the presence of ametantrone
were lower (about 2 °C) than those found for mitoxantrone,
which is consistent with its lower binding affinity (see
Table 1). Regarding the three ametantrone-containing ligands
selected in our study, all of them had a positive effect on the
thermal stability of all RNAs, particularly when the anthra-
quinone derivative was combined with neamine. This stabiliz-
ation is substantially higher than that obtained with neamine
or ametantrone alone (see Table 2), which supports the partici-
pation of both moieties in the interaction with the target.
Interestingly, the stabilizing effect of the ametantrone–
neamine ligands was considerably higher than that of mitox-
Fig. 6 Comparison of the stabilizing effect of the ligands upon com-
plexation with +3 mutated Tau RNA at different NaCl concentrations.
Melting temperatures (T
phate buffer, pH 6.8, 0.1 mM Na
m
) were determined in 10 mM sodium phos-
EDTA and NaCl at the concentration
2
indicated, at equimolecular concentrations of RNA and ligand (1 µM).
ametantrone). As expected, the stability of free RNA signifi-
antrone (e.g. the ΔT
m
values for the +3 and +14 mutated
+
cantly increased with the Na concentration (T
m
= 44.3, 51.4 or
sequences in the presence of 1 mol equivalent of mitoxantrone
were +5.7 °C and +4.6 °C, respectively, whereas in the presence
52.5 °C in 50, 100 or 150 mM NaCl, respectively). However,
as shown in Fig. 6, the stability of the complexes clearly
of the same amount of Amt–Nea2, the ΔT
m
values were
+
decreased upon increasing the Na concentration, which con-
+
12.9 °C and +10.4 °C, respectively). Moreover, the RNA stabil-
firms the importance of the electrostatic interactions between
these ligands and RNA. Interestingly, this behaviour was also
found with ametantrone, which indicates that the stabilization
provided by this anthraquinone derivative is not only based on
intercalation but also on electrostatic interactions. This is in
good agreement with the structure of the complex Mtx–Tau
RNA and with the NMR data on the interaction between
Amt–Nea2 and Tau RNA reported in this work that reveals
specific contacts between the amino functions of the side
chains of the anthraquinone heteroaromatic fragment and
RNA.
ization caused by ametantrone–neamine ligands was much
higher than that of our previous ligands based on the combi-
1
3a
nation of acridine with neamine.
value of wt Tau RNA was clearly increased in the presence
of the ametantrone–neamine ligand with the shortest spacer
ΔTm = +8.9 °C for Amt–Nea2 vs. ΔTm = +6.4 °C for Amt–Nea).
As shown in Table 2, the
T
m
(
1
1
This tendency was also reproduced in the case of the mutated
sequences and is in good agreement with previous results with
acridine–neamine ligands where a correlation between the
ligand’s binding affinity for Tau mutants and thermal stabiliz-
1
3a
ation was found.
Since the stability of RNA depends on the ionic strength
and the most promising ligands (Amt–Nea/Nea2) are positively
charged at physiological pH, we examined the influence of
sodium ion concentration on the thermal stability of the target
Conclusions
2
1
in the presence of these compounds. Melting curves of the In this work, we have described a new family of Tau RNA
most destabilized sequence, the +3 mutated RNA, were ligands based on the derivatization of one of the side chains
recorded at two NaCl concentrations (50 mM and 150 mM) in of ametantrone with azaquinolone (Amt–Azq) or neamine
the absence and in the presence of either ametantrone– (Amt–Nea/Nea2). All the reported compounds bind Tau
neamine ligands or their isolated components (neamine or RNA with low EC50 values (on the nM scale), particularly those
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
containing the aminoglycoside moiety (e.g. EC50 = 70.6 nM for
NMR spectra were recorded at 25 °C on Varian spectro-
Amt–Nea2). In addition, ametantrone-containing ligands are meters (500 or 600 MHz) using deuterated solvents. The
able to stabilize both wt Tau RNA and some of the mutated residual signal of the solvent (CD OD or D O) was used as a
3
2
1
13
sequences associated with the development of tauopathies, reference in H and C spectra. Chemical shifts are reported
such as FTDP-17. The overall data indicate that the derivatiza- in parts per million (ppm) on the δ scale, coupling constants
tion of one of the side chains of ametantrone with neamine in Hz and multiplicity as follows: s (singlet), d (doublet),
through a short spacer results in the compound with the best t (triplet), q (quadruplet), qt (quintuplet), m (multiplet),
RNA-binding properties. This is consistent with the results dd (doublet of doublets), td (doublet of triplets), ddd (doublet
from dynamic combinatorial chemistry experiments, and sup- of doublet of doublets), br (broad signal).
ports the choice of anthraquinone derivatives, such as ametan-
High-resolution matrix-assisted laser desorption ionization
trone, as a starting point to synthesize ligands with improved time-of-flight (MALDI-TOF) mass spectra were recorded on a
Tau RNA-binding properties. It is particularly relevant to note 4800 Plus spectrometer (ABSciex) both in positive (2,4-dihydro-
the ability of Amt–Nea2 to restore the thermodynamic stability xybenzoic acid matrix) and negative modes (2,4,6-trihydroxy-
of the mutated sequences upon complexation to a similar or acetophenone matrix with ammonium citrate as an additive).
even higher level than that of wt RNA. Moreover, NMR spectro- Electrospray ionization mass spectra (ESI-MS) were recorded
scopic studies revealed the existence of a preferred binding on a Micromass ZQ instrument with a single quadrupole
site for this ligand to Tau RNA in which the anthraquinone detector coupled to an HPLC, and high-resolution (HR) ESI-MS
fragment intercalates in the bulged region of the stem–loop on an Agilent 1100 LC/MS-TOF instrument.
structure, and the neamine moiety and the Amt side chains
interact with the major groove of the RNA stem.
Oligoribonucleotides were synthesized on an ABI 3400 DNA
automatic synthesizer (on the 1 μmol scale) using 2′-O-tert-
Current studies are underway to increase the cellular stabi- butyldimethylsilyl protection and following standard pro-
lity of the most promising ligand by replacing the disulfide cedures (phosphite triester approach). RNA, biotin and fluore-
linkage with a non-reversible isoster bond, as well as for eluci- scein phosphoramidites, solid-supports, reagents and solvents
dating by NMR spectroscopy the tridimensional structure of for oligoribonucleotide synthesis were purchased from Glen
the complex between Tau RNA and Amt–Nea2. All these Research or Link Technologies. RNase-free reagents, solutions
studies and the present work reported here will provide valu- and materials were used when manipulating oligoribonucleo-
able information for designing new ametantrone-containing tides. RNase-free water was obtained directly from a Milli-Q
ligands with improved RNA-binding properties, not only for system equipped with a 5000 Da ultrafiltration cartridge.
modulating Tau pre-mRNA splicing but also for binding to
other therapeutically relevant RNA secondary structures that cation of the oligoribonucleotides using linear gradients of
contain a characteristic bulged region where an adenine 0.1 M aqueous NH HCO , and a 1 : 1 mixture of 0.1 M aqueous
nucleotide is flanked by two G : C base pairs, such as the Rev NH HCO and ACN. A Kromasil C column (250 × 4.6 mm,
Reversed-phase HPLC was used for the analysis and purifi-
4
3
4
3
18
1
a,22
−1
response element (RRE) RNA.
10 μm, flow rate: 1 mL min ) was used for RNA analysis
whereas a semipreparative Jupiter C18 column (250 × 10 mm,
00 Å 10 μm, flow rate: 3 mL min− ) was used for purification
1
3
on a large scale. Characterization was carried out by high
resolution MALDI-TOF MS (negative mode, THAP matrix with
ammonium citrate).
Experimental
Materials and methods
Unless otherwise stated, common chemicals and solvents
The following oligoribonucleotide sequences were syn-
(HPLC grade or reagent grade quality) were purchased from thesized (* label denotes that the ends of the chains
commercial sources and were used without further purifi- were modified with 2′-O-methylribonucleosides): wt RNA:
cation. Aluminium plates coated with a 0.2 mm thick layer of 5′rG*C*GGCAGUGUGAGUACCUUCACACGUCC*C*; +3 mutated
silica gel 60 F254 (Merck) were used for thin-layer chromato- RNA: 5′rG*C*GGCAGUGUA
graphy analyses (TLC), whereas flash column chromatography mutated RNA: 5′rG*C*GGCAGUGUGAGUACCUUCAU
purification was carried out using silica gel 60 (230–400 C*C*. +16 mutated RNA: 5′rG*C*GGCAGUGUGAGUACCUUCA-
mesh). CAUGUCC*C*. In fluorescence binding assays, fluorescein was
̲
̲
̲
Reversed-phase high-performance liquid chromatography attached at the 5′-end of wt RNA. The following wt RNA
HPLC) analyses of thiol-derivatized monomers and ligands sequence was used in NMR studies: 5′rGGCAGUGUGAG-
(
were carried out on a Jupiter Proteo C18 column (250 × UACCUUCACACGUC.
−
1
4
.6 mm, 90 Å 4 μm, flow rate: 1 mL min ) using linear gradi-
Synthesis of the thiol-derivatized ametantrone monomer
ents of 0.045% TFA in H O (A) and 0.036% TFA in ACN (B). In
some cases, purification was carried out using the same
analytical column. A semipreparative Jupiter Proteo column
2
(
Amt–SH)
1,4-Bis((2-((2-hydroxyethyl)amino)ethyl)amino)anthraquinone
was used for the purification of some compounds (250 × (ametantrone or Amt). 1,4-Difluoroanthraquinone (850 mg,
−
1
1
0 mm, 90 Å 10 μm, flow rate: 3 mL min ), using linear gradi- 3.48 mmol) and N-(2-hydroxyethyl)ethylenediamine (3.5 mL,
ents of 0.1% TFA in H O (A) and 0.1% TFA in ACN (B).
34.8 mmol) were dissolved in dry pyridine (15 mL). After
2
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
stirring for 72 h at RT under Ar, the reaction mixture, which impurified with triphenylphosphine oxide. Purification was
acquired a deep blue colour, was evaporated in vacuo to finally accomplished by MPLC eluting with a gradient from
dryness. The residue was dissolved in water (50 mL) and 0 to 100% of B (A: 0.1% TFA in H O and B: 0.1% TFA in H O–
2 2
washed with diethyl ether (2 × 50 mL). The aqueous phase was ACN 3 : 7, 600 mL each solvent). Pure fractions by MS-HPLC
taken up and extracted several times with 20% MeOH–CHCl3 (linear gradient from 0 to 100% B in 18 min; A: 0.1% formic
until the blue colour disappeared (approx. 5 × 50 mL). Finally, acid in H
the resulting organic phase was dried over anhydrous MgSO were combined and lyophilized, providing the desired product
filtered and concentrated in vacuo to dryness to afford the as a blue solid (79 mg, 36%). Rf (AcOEt): 0.68; H NMR
crude of the expected compound, which was used without (500 MHz, CD OD) δ (ppm): 8.28 (2H, H , H , m), 7.72 (2H,
further purification in the following step. R (CHCl –MeOH , H , m), 7.54 (1H, H /H , m), 7.47 (1H, H /H , m), 3.67
2 t
O and B: 0.1% formic acid in ACN; R = 19.4 min)
4
,
1
3
G
F
f
3
H
E
H
B
C
B
C
1
7
: 3): 0.41; H NMR (500 MHz, D O) δ (ppm): 8.04 (2H, H , H , (10H, H , H , H , H , H , m), 3.39 (4H, H , H , m), 3.03 (m,
2
G
F
K
K′
M
M′
N′
L
L′
dd, J = 6.0 Hz, J = 2.9 Hz), 7.81 (2H, H
E
, H
, s), 3.93 (4H, H
.79 (4H, H , H , t, J = 6.0 Hz), 3.43 (4H, H , H , t, J = 6.0 Hz), CD OD)
H
, ddd, J = 6.0 Hz, J = 2H, H
N
), 2.29 (3H, H
/HO′, H
(ppm): 197.10, 183.25, 157.44, 147.66,
M
, HM′, m); MALDI-TOF MS, positive mode: m/z 135.73, 133.24, 127.06, 125.03, 110.93, 81.71, 61.39, 51.83,
R
, s), 1.46 (9H, H
O
/HO′, H
P Q Q
/HP′, H /H , s),
1
3
3
3
3
4
.4 Hz, J = 0.8 Hz), 7.04 (2H, H
B
, H
C
N
, HN′, m), 1.35 (9H, H
O
P
/HP′, H /H
Q
Q
, s); C NMR (125 MHz,
δ
K
K′
L
L′
3
.34 (2H, H
13.5 (calcd mass for C22
+
29 4 4
H N O [M + H] : 413.22), 435.6 47.98, 41.53, 30.71, 29.13, 27.16; MALDI-TOF MS, positive
+
+
+
[
M + Na] (calcd mass for C H N NaO [M + Na] : 435.20).
mode: m/z 671.3103 (calcd mass for C H N O S [M + H] :
34 47 4 8
2
2
28
4
4
1
46 4 8
,4-Bis((2-((2-hydroxyethyl)-N-Boc-amino)ethyl)amino)anthra- 671.3109), m/z 693.2925 (calcd mass for C34H N NaO S
+
quinone (1). A solution of Boc
v/v) mixture of dioxane–water (25 mL) was added to a solution
of the crude of Amt and Na CO (149.1 mg, 1.41 mmol) in a ethyl)amino)ethyl)amino)anthraquinone
: 1 (v/v) mixture of dioxane–water (75 mL). After stirring for stirred solution of thioester derivative 2 (40 mg, 0.060 mmol)
7 h at RT under Ar, the reaction mixture was evaporated in MeOH (0.5 mL) was added sodium methoxide (240 µL, 1 M
2
O (1.47 g, 6.75 mmol) in a 2 : 1 [M + Na] : 693.2929).
1-((2-((2-Hydroxyethyl)amino)ethyl)amino)-4-((2-((2-mercapto-
(Amt–SH). To a
(
2
3
2
1
in vacuo, and the residue was dissolved in ethyl acetate in MeOH, 0.24 mmol) and the mixture was stirred for 1.5 h at
(
(
75 mL). The resulting organic phase was washed with water RT under argon. Solvent was removed in vacuo and the result-
3 × 75 mL), dried over anhydrous MgSO , filtered and concen- ing crude was dissolved in a TFA–TIS–H O mixture (10 mL,
4
2
trated in vacuo to dryness. After flash column chromatography 95 : 2.5 : 2.5) and allowed to stand for 30 min at RT. After evap-
gradient: 0–100% AcOEt–hexanes; 0–2.5% MeOH–AcOEt), the oration in vacuo, the residue was dissolved in Milli-Q H
desired product was obtained as a blue solid (830 mg, 39% (10 mL) and 1,4-dithiothreitol (173 mg, 1.124 mmol) was
(
2
O
1
(
2 steps)). R
CD OD) δ (ppm): 8.28 (2H, H
.56 (1H, H /H , s), 7.46 (1H, H /H , m), 3.68 (8H, H , H , at 37 °C under Ar, the aqueous phase was washed with AcOEt
f
(1% MeOH in AcOEt): 0.23; H NMR (500 MHz, added to the aqueous phase to reduce the disulfide bond of
3
G
, H , m), 7.72 (2H, H , H , m), the resulting homodimeric compound. After stirring for 24 h
F
E
H
7
B
C
B
C
N
N′
H
K
, HK′, m), 3.60 (4H, H
/HO′, H /HP′, H /HQ′, s), 1.36 (9H, H
H , s); C NMR (100 MHz, CD OD) δ (ppm): 183, 157.43, 0 to 100% of B (A: 0.1% TFA in H O and B: 0.1% TFA in H O–
L
, HL′, m), 3.37 (4H, H
M
, HM′, m), 1.45 (6 × 10 mL) to remove 1,4-dithiothreitol and lyophilized. Purifi-
(9H, H
O
P
Q
O
/HO′, H
P
/HP′, H cation was carried out by MPLC eluting with a gradient from
Q
/
1
3
Q
3
2
2
1
6
47.70, 135.69, 133.19, 133.18, 127.03, 125.05, 110.84, 81.96, ACN 3 : 7, 600 mL of each solvent). Pure fractions by analytical
1.39, 51.43, 49.35, 41.71, 28.73; HR-ESI MS, positive mode: HPLC (linear gradient from 0 to 70% B in 30 min; A: 0.045%
+
m/z 613.3231 (calcd mass for C H N O [M + H] : 613.3232), TFA in H O and B: 0.036% TFA in ACN; R = 17.9 min) were
3
2
45
4
8
2
t
+
+
6
6
35.3062 [M + Na] (calcd mass for C32
35.3051).
H
44
N
4
NaO
8
[M + Na] : combined and lyophilized, providing the TFA salt of the
desired product as a blue solid (4.6 mg, 12%). MALDI-TOF MS,
1
-((2-((2-Hydroxyethyl)-N-Boc-amino)ethyl)amino)-4-((2-((2- positive mode: m/z 429.4 (calcd mass for C H N O S
2
2
29
4
3
+
(
(
thioacetyl)ethyl)-N-Boc-amino)ethyl)amino)anthraquinone [M + H] : 429.20), m/z 451.4 (calcd mass for C22
2). To a solution of PPh (17 mg, 0.65 mmol) in anhydrous [M + Na] : 451.18); ESI MS, positive mode: m/z 429.03 [M + H] ,
3
THF (2 mL) under Ar, DIAD (129 μL, 0.65 mmol) was added at m/z 451.08 [M + Na] .
H
28
N
4
NaO
3
S
+
+
+
0
°C. The formation of a white precipitate was observed after
stirring for 2 h at this temperature. At this point, a solution of
(200 mg, 0.33 mmol) in anhydrous THF (6 mL) was added
Synthesis of ametantrone-containing ligands
1
Amt–Nea2. 2,2′-Dithiobis(5-nitropyridine) (6.7 mg, 21.6 µmol)
dropwise. After stirring for 10 min at 0 °C, thiolacetic acid and Nea2–SH (1.7 µmol) were reacted in a 2 : 1 (v/v) mixture of
48 μL, 0.65 mmol) was added and the reaction mixture was THF–aqueous 0.1% TFA (3 mL) under argon at RT. After 17 h,
stirred overnight at RT. The reaction mixture was concentrated THF was evaporated in vacuo and the remaining blue solution
in vacuo to dryness, dissolved in AcOEt (50 mL) and washed was diluted with H O (2 mL). The aqueous phase was extracted
with water (3 × 50 mL). The combined organic phases were with Et O (5 × 5 mL) to remove the excess of 2,2′-dithiobis-
(
2
2
dried over anhydrous MgSO4, filtered and concentrated (5-nitropyridine). Then, a solution of Amt–SH (1.4 µmol) in
in vacuo to dryness. After flash column chromatography aqueous 0.1% TFA (3 mL) was added and the mixture was
(
gradient: 0–100% AcOEt–hexanes; 0–5% MeOH–AcOEt), the stirred overnight under Ar at RT. After purification by semi-
desired product was obtained as a blue solid that was slightly preparativereversed-phaseHPLC(lineargradientfrom0to35%B
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
+
in 30 min; A: 0.1% TFA in H
2
46 8 5 2
O and B: 0.1% TFA in ACN; flow: for C37H N NaO S [M + Na] : 769.2930); analytical HPLC
mL min− ; R = 23.7 min) and lyophilization, the desired (0 to 50% B in 30 min; R = 21.2 min).
1
3
t
t
1
product was obtained as a blue solid (0.78 µmol, 56%).
NMR (600 MHz, D O) δ (ppm): 8.25 (2H, H , H
H
General procedure for the RNA-templated DCC experiments
2
G
F
, m), 7.81 (2H,
1
3
H , H , m), 7.49 (2H, H , H , s), 5.46 (1H, H , m), 4.02 (1H, DCC experiments were carried out as previously described.
E
H
B
C
1′
H
a
, m), 3.86 (5H, HN′, H5′, H
6
, Ha′, m), 3.80 (4H, H
K
, HK′, m), Briefly, biotinylated wt RNA (6 nmol) was annealed in 240 µL
, HM′, H of buffer (50 mM Tris-HCl, pH 7.7, 100 mM NaCl and 0.1 mM
m), 3.33 (1H, H , m), 3.21 (4H, H , H , m), 3.09 (2H, H , H , Na EDTA) by heating to 90 °C for 5 min and then slowly
3
.72 (1H, H3′, m), 3.51 (2H, H6′, m), 3.38 (6H, H
4
, H
5
M
,
4
′
L
L′
1
3
2
m), 3.02 (1H, H2′, m), 3.00 (2H, H
N b
, m), 2.82 (2H, H , m), 2.15 cooling to room temperature. After overnight incubation at RT,
(
1H, H2,eq, m), 1.48 (1H, H2,ax, m); HR-ESI MS, positive mode: the solutions were stored at 4 °C. The annealed biotinylated
+
m/z 809.3671 (calcd mass for C H N O S [M + H] : RNA was added to an Eppendorf tube containing the quanti-
3
6
57 8 9 2
8
[
09.3690), m/z 831.3497 (calcd mass for C36
H
56
N
8
NaO
M + Na] : 831.3509); analytical HPLC (0 to 35% B in 30 min: monomer), and the resulting mixture was left to stand at RT
Rt = 21.6 min). under air without stirring. At the desired time, the disulfide
9
S
2
fied (Ellman’s test) thiol building blocks (24 nmol of each
+
Amt–Nea. 2,2′-Dithiobis(5-nitropyridine)(11.6mg, 37.5µmol) exchange was stopped by the addition of aqueous 0.1% TFA
and Nea–SH (3 µmol) were reacted in a 2 : 1 (v/v) mixture of (v/v) solution until pH ∼5–6 was reached.
THF–aqueous 0.1% TFA (3 mL) under argon for 17 h at RT.
The biotinylated RNA and the binding ligands were isolated
After evaporation in vacuo and washing with Et
2
O, a solution with streptavidin-coated magnetic beads (Biomag Streptavidin,
of Amt–SH (1.7 µmol) in aqueous 0.1% TFA (3 mL) was added 5 mg mL− suspension, Qiagen). A magnet was used in all
and the mixture was stirred overnight under argon at RT. After washing procedures to retain the beads in the tube while the
purification by HPLC (see the conditions used for Amt–Nea2) supernatant was pipetted off. The beads (500 µL of suspension
1
and lyophilization, the desired product was obtained as a blue for each DCL aliquot) were first washed with an acidic buffer
1
solid (0.57 µmol, 33%). H NMR (600 MHz, D O) δ (ppm): 8.32 (3 × 500 µL of 50 mM Tris-HCl, pH 5.8, 100 mM NaCl and
2
(2H, H
G F E H B C 2
, H , m), 7.87 (2H, H , H , m), 7.57 (2H, H , H , m), 0.1 mM Na EDTA). After incubation for 20 min at RT with the
5
.42 (1H, H1′, d, J = 3.7 Hz), 3.95 (3H, H5′, HN′, m), 3.85 (4H, DCL aliquots, the beads were retained in the vessel using the
H , H , m), 3.78 (1H, H , m), 3.66 (1H, H ), 3.59 (1H, H , m), magnet and the supernatant solution was pipetted off again.
K
K′
a
3′
a′
3
3
2
.50 (1H, H
6
), 3.45 (3H, H
, HL′, m), 3.18 (2H, HM′, m), 2.99 (2H, H
M
, H
4
), 3.39 (4H, H
5
, H6′, H4′, m), After washing the beads (3 × 200 µL of 50 mM Tris-HCl, pH
.24 (4H, H
L
1
, H
3
, m), 5.8, 100 mM NaCl and 0.1 mM Na
2
EDTA) to remove the non-
.87 (3H, H , H , m), 2.50 (2H, H , t, J = 7.1 Hz), 2.16 (1H, H
interacting compounds, a treatment with a hot solution of
2
′
N
f
2,
eq, m), 1.43 (5H, H
b
, H
e
, H2,ax, m), 1.10 (4H, H
c
, H
d
2
); HR-ESI 0.1% TFA in H O (v/v) was used to liberate RNA-binding
MS, positive mode: m/z 865.4324 (calcd mass for C40
[
H
65
N
8
O
9
S
2
ligands (3 × 200 µL, incubation at 90 °C for 10 min). The three
M + H] : 865.4316); analytical HPLC (0 to 35% B in 30 min; acidic solutions were combined and evaporated to dryness in a
= 23.8 min). Speed-Vac to provide a residue that was dissolved in 0.1% TFA
Amt–Azq. 2,2′-Dithiobis(5-nitropyridine) (2.4 mg, 7.5 µmol) in H O (v/v) and analysed by UV-MS HPLC. Elution was per-
and Azq-SH (1.5 µmol) were reacted in a 1 : 1 (v/v) mixture of formed on a Kinetex C column (100 × 4.6 mm, 2.6 μm, 100 Å,
+
R
t
2
1
8
THF–1 M aqueous ammonium acetate pH = 3.5 (4 mL) under flow rate: 0.2 mL min− ) with linear gradients of H
1
O and ACN
2
argon at RT. After 5 h, THF was evaporated in vacuo and the containing both solvents, either 0.1% formic acid (v/v) or 0.1%
remaining solution was diluted with H O (2 mL). The aqueous formic acid (v/v) and 0.01% TFA (v/v), to avoid the overlapping
2
phase was extracted with Et
A solution of Amt–SH in H O–0.1% TFA (2 mL) was added to peak areas of the HPLC traces were integrated and normalized
the activated azaquinolone monomer dissolved in 1 M taking into account the previously described extinction coeffi-
2
O (5 × 5 mL) and lyophilized. of some peaks and to allow a more accurate integration. All
2
1
3a
aqueous ammonium acetate pH = 3.8 (3 mL) under Ar. After cients at 260 nm.
The following extinction coefficient
stirring for 3 h at RT, Amt–Azq ligand was isolated by semi- was calculated for ametantrone-containing compounds: ε260
:
preparative reversed-phase HPLC (linear gradient from 0 to 39.253 M− cm
0% B in 30 min; A: 0.1% formic acid in H O and B: 0.1%
= 16.2 min) and after
lyophilization, the desired product was obtained as a blue All the assayed compounds displayed a purity ≥95%, deter-
1
−1
.
4
2
−1
Evaluation of the interaction between RNA and ligands
formic acid in ACN; flow: 3 mL min ; R
t
1
solid (0.73 µmol, 48%). H NMR (600 MHz, D
1H, H /H , m), 7.94 (1H, H /H , m), 7.75 (2H, H
.56 (2H, H , H , m), 7.22 (2H, H , H ), 6.79 (1H, H , d, J = recorded as previously described.
2
O) δ (ppm): 7.98 mined by HPLC analysis.
(
7
8
F
G
G
F
E
, H , m), UV-monitored melting experiments. Melting curves were
H
1
3a,b
Briefly, samples were
W
V
B
C
U
−
1
.0 Hz), 6.38 (1H, H
X
, d, J = 9.5 Hz), 3.86 (4H, HN′, H
/HK′, m), 3.31 (4H, H
T
, m), 3.76 cooled from 90 °C to 20 °C at a constant rate of 0.5 °C min
/HL′, H and the absorbance at 260 nm was measured as a function of
(
2H, H
K
/HK′, m), 3.64 (2H, H
K
L
Q
,
m), 3.19 (4H, H /H , H , m), 3.12 (2H, HM′, m), 3.00 (4H, HO, temperature. The denaturation curves (20 °C to 90 °C) were
L
L′
M
H
N
, m), 2.93 (2H, H
S
, m), 2.68 (2H, H
P
, m), 1.86 (2H, H
R
, m); also recorded. All experiments were repeated at least three
values were obtained. The estimated
C H N O S [M + H] : 747.3111), m/z 769.2945 (calcd mass error in Tm values was ±0.2 °C. The solutions were 1 µM both
HR-ESI MS, positive mode: m/z 747.3119 (calcd mass for times until coincident T
m
+
3
7
47 8 5 2
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
in RNA (wt, +3, +14 or +16) and in ligands, in 10 mM sodium processed with Topspin software. Samples were dissolved in
phosphate buffer pH 6.8, 100 mM NaCl and 0.1 mM Na EDTA. 10 mM phosphate buffer, pH 6.8, in either D O or H O–D O
2
2
2
2
UV-Vis and fluorescence titration experiments. First, UV-Vis 9 : 1.
spectra or fluorescence emission spectra of the free ligand in
For titration experiments, wt RNA (60 nmol) was dissolved
the previous buffer were recorded. Subsequent aliquots of a wt in 600 µL of 10 mM sodium phosphate buffer (pH 6.8), and
RNA solution in the same buffer (previously folded by heating 2 µL of a 33.3 µM solution of DSS in water were added. After
the solution for 5 min to 90 °C and left to slowly cool to RT) lyophilization, the residue was dissolved in 600 µL of a 9 : 1
were added to the ligand’s solution, and the UV-Vis or fluore- H O–D O mixture. Annealing was performed by heating at
2
2
scence emission spectra were recorded.
90 °C for 3 min, followed by snap cooling on ice for 20 min.
Fluorescence binding assays. Fluorescence measurements NMR spectra were recorded at 5 °C to minimize the exchange
were performed in 1 cm path-length quartz cells on a Quanta- with water. Water suppression was achieved by using an exci-
1
Master fluorometer (PTI) at 25 °C, with an excitation slit width tation sculpting sequence (zgesgp). Subsequently, 1D H NMR
of 4.8 nm and an emission slit width of 8 nm. Upon excitation spectra were recorded at 5 °C on the addition of increasing
at 490 nm, the emission spectrum was recorded over a range quantities of the ligand (0.5, 1.0, 1.5 and 2.0 mol equiv.). The
between 505 and 540 nm until no changes in the fluorescence total volume of the sample never changed more than 5%.
intensity were detected. All binding assays were performed in
A 0.28 mM RNA sample (147 nmol in 525 µL sodium phos-
the melting curves’ buffer. The sample was continuously phate buffer) was used for recording 2D spectra in H O–D O at
2
2
stirred except during the fluorescence measurement. First, the 5 °C. The complex with Amt–Nea2 was prepared by adding
fluorescence spectrum of the buffer was recorded, to be used 1 mol equivalent of the ligand. A series of DQF-COSY, TOCSY
as the baseline. For each experiment, the spectrum of a 25 nM and NOESY experiments was recorded in D O. The NOESY
2
solution of refolded fluorescein-derivatized RNA (25 pmol in spectra were acquired with mixing times of 100, 250 and
1
mL buffer) was recorded, and the baseline blank subtracted. 300 ms, and the TOCSY spectra were recorded with the stan-
Subsequent aliquots of 1 µL of aqueous ligand’ solutions (of dard MLEV-17 spin-lock sequence and 80 ms mixing time. In
increasing concentrations, depending on the ligand’s affinity) addition, NOESY spectra in H O were acquired with 100 ms
were added to the RNA-containing solution. The fluorescence mixing time. In 2D experiments in H O, water suppression was
2
2
spectrum was recorded after addition of each aliquot until the achieved either by using an excitation sculpting sequence or
fluorescein fluorescence signal at 517 nm reached saturation by including a WATERGATE module in the pulse sequence
(
typically 5–10 min). Over the entire range of ligand concen- prior to acquisition. Two-dimensional experiments in D
trations used, the emission maxima varied less than 1 nm. The were carried out at temperatures ranging from 5 °C to 25 °C,
total volume of the sample never changed more than 10–15%. whereas spectra in H O were recorded at 5 °C to reduce the
2
O
2
The full titration was repeated in the absence of labelled RNA exchange with water. The spectral analysis program Sparky was
to correct for the presence of the ligand’s fluorescence. These used for semiautomatic assignment of the NOESY cross-peaks
spectra were subtracted from each corresponding point of the and quantitative evaluation of the NOE intensities. Structural
labelled RNA titrations, and the resulting fluorescence inten- models based on NMR constraints were built with the program
1
1
sity was corrected for dilution (F′ = F × V/V0).
Sybyl (Tripos Inc.) on the basis of the Mtx : Tau RNA complex
The emission fluorescence at 517 nm was normalized by (PDB code: 1Qc8). The ametantrone and neamine moieties
dividing the difference between the observed fluorescence, F′, were set at similar positions as the mitoxantrone and neo-
9
,11
and the final fluorescence, F , by the difference between the mycin in their respective complexes with the RNA.
Initial
f
initial fluorescence, F , and the final fluorescence, F . This nor- coordinates of the complex were energy minimized, submitted
0
f
malized fluorescence intensity was plotted as a function of the to a short run of molecular dynamics, and energy minimized
logarithm of the total ligand concentration in each point of again to optimize the geometry and eliminate steric contacts.
the titration. Finally, nonlinear regression using a sigmoidal These calculations were carried out with the program Sybyl
dose–response curve was performed with the software package using built-in force fields. Analysis of the structures was
GraphPad Prism 4 (GraphPad Software, San Diego, CA) to carried out with the programs Sybyl and MOLMOL.
calculate the EC50 values. Experimental errors were less than
or equal to ±25% of each value.
For competitive experiments, a tRNA from baker’s yeast
S. cerevisiae) was purchased from Sigma. Stock solutions of
Acknowledgements
(
mix
tRNA
of 9.640 cm− per base.
were carried out as described above with the exception that a and L. Ortiz from the facilities of the Servei d’Espectrometria
were quantified using an average extinction coefficient The authors acknowledge Dr M. Gairí from the Barcelona
1
13a
The fluorescence binding assays Scientific Park for NMR technical support and Dr I. Fernández
mix
3
0-fold excess (base) of the tRNA
fluorescein-labelled RNA (or to the buffer for the titration work was supported by funds from the Spanish Ministerio de
without target RNA). Ciencia e Innovación (grant CTQ2010-21567-C02-01-02, the
was added to the refolded de Masses of the University of Barcelona for MS support. This
NMR spectroscopy. NMR spectra were acquired on Bruker Generalitat de Catalunya (2009SGR-208) and the Xarxa de
Advance spectrometers operating at 600 or 800 MHz, and Referència de Biotecnologia) and the Programa d’Intensifica-
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
ció de la Recerca (Universitat de Barcelona). Paula López-Senín
received a fellowship from the Ministerio de Educación y
Ciencia and Gerard Artigas from the Universitat de Barcelona.
9 L. Varani, M. G. Spillantini, M. Goedert and G. Varani,
Nucleic Acids Res., 2000, 28, 710–719.
10 C. P. Donahue, J. Ni, E. Rozners, M. A. Glicksman and
M. S. Wolfe, J. Biomol. Screen., 2007, 12, 789–799.
11 S. Zheng, Y. Chen, C. P. Donahue, M. S. Wolfe and
G. Varani, Chem. Biol., 2009, 16, 557–566.
Notes and references
1
2 Y. Liu, E. Peacey, J. Dickson, C. P. Donahue, S. Zheng,
G. Varani and M. S. Wolfe, J. Med. Chem., 2009, 52, 6523–
6526.
1
(a) J. Gallego and G. Varani, Acc. Chem. Res., 2001, 34,
36–843; (b) Y. Tor, ChemBioChem, 2003, 4, 998–1007;
c) P. A. Sharp, Cell, 2009, 136,
d) W. E. Georgianna and D. D. Young, Org. Biomol. Chem.,
011, 9, 7969–7978.
(a) C. S. Chow and F. M. Bogdan, Chem. Rev., 1997, 97,
489–1513; (b) G. J. R. Zaman, P. J. A. Michiels and
C. A. A. van Boeckel, Drug Discovery Today, 2003, 8, 297–
06.
(a) J. R. Thomas and P. J. Hergenrother, Chem. Rev., 2008,
8
(
(
577–580; 13 (a) P. López-Senín, I. Gómez-Pinto, A. Grandas and
V. Marchán, Chem. Eur. J., 2011, 17, 1946–1953;
–
2
(b) P. López-Senín, G. Artigas and V. Marchán, Org. Biomol.
Chem., 2012, 10, 9243–9254; (c) P. López-Senín, Ph.D.
Thesis, University of Barcelona, Spain, 2010; (d) G. Artigas
and V. Marchán, J. Org. Chem., 2013, 78, 10666–10677.
14 (a) P. T. Corbett, J. Leclaire, L. Vial, K. R. West, J.-L. Wietor,
J. K. M. Sanders and S. Otto, Chem. Rev., 2006, 106, 3652–
3711; (b) Dynamic Combinatorial Chemistry: In Drug Discov-
ery, Bioorganic Chemistry, and Materials Science, ed.
B. L. Miller, John Wiley & Sons, Inc., Hoboken, N. J., 2010;
(c) C. Sherman Durai and M. M. Harding, Aust. J. Chem.,
2011, 64, 671–680; (d) A. Herrmann, Chem. Soc. Rev., 2014,
43, 1899–1933; (e) B. D. Blakeley, S. M. DePorter, U. Mohan,
R. Burai, B. S. Tolbert and B. R. McNaughton, Tetrahedron,
2012, 68, 8837–8855.
2
3
1
3
108, 1171–1224; (b) F. Aboul-ela, Future Med. Chem., 2010,
2, 93–119; (c) L. Guan and M. D. Disney, ACS Chem. Biol.,
2012, 7, 73–86; (d) L. O. Ofori, J. Hoskins, M. Nakamori,
C. A. Thornton and B. L. Miller, Nucleic Acids Res., 2012,
0, 6380–6390; (e) J. L. Childs-Disney, R. Parkesh,
4
M. Nakamori, C. A. Thornton and M. D. Disney, ACS Chem.
Biol., 2012, 7, 1984–1993; (f) L. Guan and M. D. Disney,
Angew. Chem., Int. Ed., 2013, 52, 1462–1465; (g) J.-P. Joly,
G. Mata, P. Eldin, L. Briant, F. Fontaine-Vive, M. Duca and 15 (a) A. Gianoncelli, S. Basili, M. Scalabrin, A. Sosic, S. Moro,
R. Benhida, Chem. – Eur. J., 2014, 20, 2071–2079.
(a) T. Xia, Curr. Opin. Chem. Biol., 2008, 12, 604–611;
G. Zagotto, M. Palumbo, N. Gresh and B. Gatto, ChemMed-
Chem, 2010, 5, 1080–1091; (b) A. Skladanowski and
J. Konopa, Br. J. Cancer, 2000, 82, 1300–1304.
4
(
(
b) T. Tuccinardi, Future Med. Chem., 2011, 3, 723–733;
c) A. C. Stelzer, A. T. Frank, J. D. Kratz, M. D. Swanson, 16 B. Stefanska, M. Dzieduszycka, S. Martelli and E. Borowski,
J. Med. Chem., 1989, 32, 1724–1728.
D. M. Markovitz and H. M. Al-Hashimi, Nat. Chem. Biol., 17 F. Rabanal, W. F. DeGrado and D. L. Dutton, Tetrahedron
011, 7, 553–559; (d) D. I. Bryson, W. Zhang,
Lett., 1996, 37, 1347–1350.
P. M. McLendon, T. M. Reineke and W. L. Santos, ACS 18 (a) P. C. Gareiss, K. Sobczak, B. R. McNaughton,
M. J. Gonzalez-Hernandez, J. Lee, I. Andricioaei,
2
Chem. Biol., 2012, 7, 210–217.
P. B. Palde, C. A. Thornton and B. L. Miller, J. Am.
Chem. Soc., 2008, 130, 16254–16261; (b) N. W. Luedtke,
Q. Liu and Y. Tor, Biochemistry, 2003, 42, 11391–11403;
(c) J. R. Thomas, X. Liu and P. J. Hergenrother, Biochemis-
try, 2006, 45, 10928–10938; (d) S. J. Lee, S. Hyun, J. S. Kieft
and J. Yu, J. Am. Chem. Soc., 2009, 131, 2224–2230.
5
6
(a) C. Ballatore, V. M. Lee and J. Trojanowski, Q. Nat. Rev.
Neurosci., 2007, 8, 663–672; (b) W. Noble, A. M. Pooler and
D. P. Hanger, Expert Opin. Drug. Discov., 2011, 6, 797–810.
(a) L. Varani, M. Hasegawa, M. G. Spillantini, M. J. Smith,
J. R. Murrell, B. Ghetti, A. Klug, M. Goedert and G. Varani,
Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 8229–8241; 19 L. Guan and M. D. Disney, Angew. Chem., Int. Ed., 2013, 54,
(
b) C. P. Donahue, C. Muratore, J. Y. Wu, K. S. Kosik and
1462–1465.
M. S. Wolfe, J. Biol. Chem., 2006, 281, 23302–23306.
(a) M. G. Spillantini, J. R. Murrell, M. Goedert, M. R. Farlow
and A. Klug, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 7737–
20 (a) R. Moumné, M. Catala, V. Larue, L. Micouin and
C. Tisné, Biochimie, 2012, 94, 1607–1619; (b) T. Lombes,
R. Moumné, V. Larue, E. Prost, M. Catala, T. Lecourt,
F. Dardel, L. Micouin and C. Tisné, Angew. Chem., Int. Ed.,
2012, 51, 9530–9534.
7
8
7
8
741; (b) F. Liu and C. X. Gong, Mol. Neurodegener., 2008, 3,
; (c) M. S. Wolfe, J. Biol. Chem., 2009, 284, 6021–6025;
(
d) M. Niblock and J.-M. Gallo, Biochem. Soc. Trans., 2012, 21 A. M. Smith, J. Kassman, K. J. Srour and A. M. Soto, Bio-
4
0, 677–680; (e) E. Peacey, L. Rodríguez, Y. Liu and
chemistry, 2011, 50, 9434–9445.
M. S. Wolfe, Nucleic Acids Res., 2012, 40, 9836–9849.
(a) E. Buratti and F. E. Baralle, Mol. Cell. Biol., 2004, 24,
22 (a) S. R. Kirk, N. W. Luedtke and Y. Tor, J. Am. Chem. Soc.,
2000, 122, 980–981; (b) J. Zhang, S. Umemoto and
K. Nakatani, J. Am. Chem. Soc., 2010, 132, 3660–3661;
(c) T. Tran and M. D. Disney, Nat. Commun., 2012, 3, 1–9.
1
0505–10514; (b) N. N. Singh and R. N. Singh, Nucleic Acids
Res., 2007, 35, 371–389.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem.