Investigation of RNA-ligand interactions by 19F NMR spectroscopy using fluorinated probes

Article abstract of DOI:10.1002/anie.201204083

Spy swap: The interaction between an unlabeled RNA and unlabeled ligands (red hexagon) can be monitored by ¹⁹F NMR spectroscopy using small fluorinated diamines (green star) as spy reporters (see scheme). This technique also enables the visualization of the conformational capture of a riboswitch by its ligand. Copyright

Full text of DOI:10.1002/anie.201204083

Angewandte
Chemie
DOI: 10.1002/anie.201204083
RNA Structures
19
Investigation of RNA–Ligand Interactions by F NMR Spectroscopy
Using Fluorinated Probes**
Thomas Lombꢀs, Roba Moumnꢁ, Valꢁry Larue, Elise Prost, Marjorie Catala, Thomas Lecourt,
Frꢁdꢁric Dardel, Laurent Micouin,* and Carine Tisnꢁ*
Ribonucleic acid (RNA) is now recognized as playing a key
role in many biological functions, and is emerging as an
important new drug target.[1] However, its therapeutic poten-
tial is still underexploited.[2] Indeed, the limited understand-
ing of the interactions between small molecules and RNA still
hampers rational drug development of RNA-targeting mol-
ecules. Among the different methods available to investigate
binding between small molecules and RNA,[3] NMR spec-
troscopy is particularly attractive as it can deliver information
on molecular interactions at the atomic level, including
conformational rearrangements that can occur before or
upon binding.[4] The dynamic nature of this interaction is
particularly important in RNA-regulated pathways.[5]
site.[8] The use of ligand-based binding-competition NMR
screening using fluorinated ligands has been described by
Dalvit for the investigation of protein–ligand interactions
(FAXS technique),[9] but has not been applied to the study
small molecules interacting with RNA. We report herein that
competitive binding of fluorinated probes can be used to
detect and quantify the interaction between unlabeled RNA
and non-fluorinated ligands and to monitor dynamic RNA
folding events (Figure 1).
Many NMR spectroscopy techniques have been devel-
oped to visualize dynamic RNA–ligand interactions, most of
them based on the observation of either the target or ligand
1H nuclei.[6] However, some difficulties can occur when
studying larger strands of RNA, as the number of detectable
signals will increase. Introducing a specific label is one way to
overcome this problem. An elegant method based on
19F NMR spectroscopy was proposed some years ago by
Micura and co-workers. Introduction of a fluorine atom at
a specific position of RNA allows local monitoring of binding
events at this site.[7] One technical difficulty with this
approach is the need to chemically modify the RNA, which
can be difficult for large RNAs or for RNAs with modified
nucleotides. Furthermore, this modification can affect RNA–
ligand interactions.
Figure 1. Principle of the displacement experiments. A) Competition
for binding to a structured RNA between a fluorinated probe (F, star)
and an RNA ligand (L, hexagon). B) Competition for binding to a bi-
stable RNA. In both cases, the fluorinated reporter is monitored by
19F NMR.
The binding of aminoglycosides with 16S23 RNA,
a 23 nucleotide hairpin that mimics the decoding A-site of
16S ribosomal RNA, was first investigated as a case study.
Although absolute KD values are difficult to compare since
experimental conditions significantly vary from one study to
another, the KD values of the most studied ligands of 16S A-
site RNA can be ordered as follows: neomycin < paromomy-
cin < neamine < paromamine, neomycin being the strongest
binder and paromamine the weakest.[10] Fluorinated com-
pound 1 being an analogue of desoxystreptamine (DOS), the
core moiety of most aminoglycosides, we first checked that it
bound to the 16S23 RNA.[11] Its binding was monitored using
both 19F and 1H NMR spectroscopy (Figure 2). A KD of 2 mm
was determined by NMR titration. As reported for DOS,[12]
compound 1 interacts with two equivalent binding sites of
16S23 RNA in a fast-exchange regime.[13]
We then conducted ligand-based binding competition
using compound 1 as a fluorinated spy probe and known
binders of 16S23 RNA (neamine 2, paromamine 3, neomycin
4) as competitors. Progressive displacement of the fluorinated
reporter could be observed by increasing the concentration of
the competitor (Figure 3), the KD value of paromamine,
neamine, and neomycin for 16S23 could be estimated in the
We have recently shown that 19F NMR spectroscopy can
be used to monitor the binding of racemic fluorinated
molecules to various RNAs and that chiral recognition can
be used to monitor the local conformation of the binding
[*] T. Lombꢀs, Dr. R. Moumnꢁ, E. Prost, Dr. T. Lecourt, Dr. L. Micouin
UMR 8638 University Paris Descartes, CNRS, Facultꢁ de Pharmacie
4 av. de l’Observatoire, 75006 Paris (France)
E-mail: laurent.micouin@parisdescartes.fr
Homepage: http://www.ssmip.cnrs.fr/eq3_micouin.html
Dr. V. Larue, M. Catala, Prof. F. Dardel, Dr. C. Tisnꢁ
UMR 8015 University Paris Descartes, CNRS, Facultꢁ de Pharmacie
4 av. de l’Observatoire, 75006 Paris (France)
E-mail: carine.tisne@parisdescartes.fr
Homepage: http://lcrbw.pharmacie.univ-paris5.fr/spip.php?rubri-
que10
[**] Financial support from CNRS Bettencourt Schueller Foundation
(PhD grant to T.L.), ANRS (grant to R.M.) and ANR (research
project PCV TriggeRNA) is acknowledged.
Supporting information for this article (experimental details) is
available on the WWW under http://dx.doi.org/10.1002/anie.
201204083.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
other using reporter 1, were better differ-
entiated using a tighter 16S23 RNA binder
like compound 10 at a high-salt concentra-
tion.[7] Thus, two fluorinated probes that
present a broad spectrum of interaction with
RNA targets can be used to design specific
ligands towards an RNA target. In a frag-
ment-based approach, compound 1 is a better
suited molecule to perform competition
experiments whereas, in optimization assays,
compound 10 will be more efficient.
We then investigated the use of fluori-
nated compounds to monitor RNA confor-
mational changes upon ligand binding (prin-
ciple from Figure 1B). As a case study, we
investigated the interaction of neomycin with
a known neomycin aptamer.[18] This aptamer
is particularly interesting since it adopts an
unbound “on” conformation and a bound
“off” conformation in the presence of its
ligand.[18] This conformational switch can be
visualized by 1H NMR, (Figure 5A,D).
Ligand binding induces extensive structural
changes, as indicated by the appearance of
new imino protons upon addition of neo-
mycin (Figure 5A).[19] We first analyzed the
interaction between the aptamer and com-
Figure 2. Measurement of the KD between compound 1 and 16S23 RNA (0.3 mm) by
19F NMR spectroscopy in KPO4 buffer pH 6.5 containing 50 mm KCl at 293 K. A) 19
chemical shift for increasing concentrations of probe. B) KD determination based on the
19F NMR chemical-shift variation (observation of the probe). C) KD determination based
on 1H NMR chemical-shift variation of the 16S23 G17 base (observation of the target).
F
range of 120 mm, 50 mm, and 35 mm, respectively. The ranking
of the KD values of these molecules is thus in agreement with
published data.[10] However, we would have expected a greater
difference in the KD values of neomycin and neamine. The
inability of compound 1 to accurately measure this difference
might be from too great a difference in dissociation constants
between reporter 1 and neomycin.[14]
For the purpose of ranking higher affinity ligands, we
prepared a novel fluorinated reporter that could bind in
a more selective and efficient manner. Compound 10 was
therefore prepared from paromomycin 5 (Scheme 1). This
compound showed a sixfold lower dissociation constant for
the 16S23 RNA than compound 1, as determined by 1H NMR
titration (KD = 300 mm).[15] However, its improved KD led to
a significant broadening of its 19F NMR signal in the presence
of 16S23 RNA, as a result of an intermediate exchange
regime. Fortunately, a sharp signal could be recovered by
increasing the ionic strength of the buffer to 300 mm KCl.[15]
Working at a high salt concentration has two interesting
consequences: 1) it decreases the KD of the reporter for the
RNA target, and the interaction is back to a fast regime
exchange on NMR timescale; 2) the electrostatic part of the
binding interaction is lowered, and thus the promiscuous
binding that often occurs with an RNA target is limited.[16]
Using this reporter 10 under these conditions, we estimated
Scheme 1. Reaction conditions: a) HCl, MeOH, heated to reflux, over-
night; b) benzylchloroformate (CbzCl), Na2CO3, toluene/water/ace-
tone, À58C to RT, 4 h; c) tert-butyldimethylsilyl chloride (TBDMSCl), 4-
dimethylaminopyridine (DMAP), Et3N, dimethylformamide (DMF),
À58C, 1 h; d) Ac2O, DMAP, pyridine, RT, 1 day; e) HF/pyridine,
tetrahydrofuran (THF), RT, 4 h; f) diethylaminosulfur trifluoride
(DAST), dry dichloroethane, À58C to RT, 5 h; g) NaOMe, MeOH/THF/
dioxane, RT, 3 h; h) H2, Pd/C, dioxane/MeOH/water, RT, 2 days.
a KD of 60 mm for neomycin, 90 mm for paromomycine and
250 mm for neamine (Figure 4). This ranking is in agreement
with ones determined by other techniques.[10] Interestingly,
raising the salt concentration increases KD values of 16S23
ligands, and tends to increase the differences between good
ligands, thus allowing a more precise ranking.[17] Thus,
neamime and neomycin, initially ranked very close to each
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Figure 4. Competition experiments with 10 as the fluorinated probe:
19F NMR spectra of compound 10 (0.7 mm) mixed with 16S23 RNA
(0.3 mm) in KPO4 buffer pH 6.5, containing 300 mm KCL, at 293 K,
with increasing concentrations of A) neamine, B) paromomycin, and
C) neomycin.
Figure 3. Competition experiments using compound 1 as the fluori-
nated probe. Right: 19F NMR spectra of compound 1 (0.7 mm) mixed
with 16S23 RNA (0.3 mm) in KPO4 buffer pH 6.5 containing 50 mm
KCl at 293 K with increasing concentrations of A) paromamine, B) nea-
mine, and C) neomycin.
pounds 1 and 10 (Figure 5). Both probes bind to the aptamer
but with different outcomes (Figure 5B,C). Compound 1 bind-
ing does not alter the conformational equilibrium of the RNA
target (Figure 5C). Only chemical-shift variations are
observed upon addition of compound 1 leading to a KD
value in the millimolar range. This result supports the idea
that this external probe will induce only very limited bias in
the study of the interaction between ligands and the RNA
target. On the contrary, the higher affinity of compound 10
for the aptamer is illustrated by a complete disappearance of
Figure 5. 1H NMR spectra of neomycin aptamer (0.2 mm) in KPO4
buffer pH 6.5 containing 50 mm KCl, at 293 K with A) one equivalent
of neomycin, B) compound 10 (0.46 mm), C) compound 1 (1.4 mm),
and D) alone.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
RNA. Chemical modification of the
reporter and tuning of experimental con-
ditions enable the observation of ligands
with different binding strengths. Fluori-
nated reporters can also be used to monitor
the conformational capture of a bi-stable
RNA upon binding. In this case, a reporter
with a low affinity to the target behaves as
a neutral partner and does not influence
the equilibrium between the bound and
unbound states, in contrast to a higher-
affinity binder. The absence of any target
or ligand labeling step, the 100% natural
abundance of the 19F nucleus, the bio-
orthogonality of the carbon–fluorine bond,
as well as the high 19F NMR sensitivity
make this method very attractive from
a practical standpoint.[20] We believe that
19F NMR spectroscopy competitive screen-
ing will be a simple and efficient method
for the discovery of new RNA binders.
Received: May 25, 2012
Published online: && &&, &&&&
Keywords: aminoglycosides ·
.
19F NMR spectroscopy · reporter ligands ·
riboswitches · RNA
Figure 6. Dynamic folding of neomycin aptamer monitored by NMR spectroscopy using the
fluorinated reporter 1. 19F NMR spectra of neomycin aptamer (0.3 mm) in a phosphate buffer
(10 mm, pH 6.5) with 50 mm KCl and A) compound 1 (0.7 mm), B) compound 1 (0.7 mm)
1
and neomycin (0.45 mm). H NMR spectra of neomycin aptamer (0.3 mm) in a phosphate
buffer (10 mm, pH 6.5) with 50 mm KCl, C) alone, D) compound 1 (0.7 mm), E) compound
1 (0.7 mm) and neomycin (0.45 mm), or F) neomycin (0.3 mm).
[1] a) P. A. Sharp, Cell 2009, 136, 577 – 580;
b) T. A. Cooper, L. Wan, G. Dreyfuss, Cell
2009, 136, 777 – 793; c) M. Esteller, Nat.
Rev. Genet. 2011, 12, 861 – 874.
its 19F NMR signal upon binding, as a result of an intermedi-
[2] L. Guan, M. D. Disney, ACS Chem. Biol. 2012, 7, 73 – 86.
[3] J. R. Thomas, P. J. Hergenrother, Chem. Rev. 2008, 108, 1171 –
1224.
[4] a) M. Pellecchia, I. Bertini, D. Cowburn, C. Dalvit, E. Giralt, W.
Jahnke, T. L. James, S. W. Homas, H. Kessler, C. Luchinat, B.
Meyer, H. Oschkinat, J. Peng, H. Schwalbe, G. Siegal, Nat. Rev.
Drug Discovery 2008, 7, 738 – 745; b) C. Dalvit, Drug Discovery
Today 2009, 14, 1051 – 1057.
[5] a) J. Almeida Cruz, E. Westhof, Cell 2009, 136, 604 – 609; b) J.
Parsons, M. P. Castaldi, S. Dutta, S. M. Dibrov, D. L. Wyles, T.
Hermann, Nat. Chem. Biol. 2009, 5, 823 – 825; c) U. Rieder, C.
Kreutz, R. Micura, Proc. Natl. Acad. Sci. USA 2010, 107, 10804 –
10809; d) K. E. Deigan, A. R. Ferrꢀ-dꢁAmarꢀ, Acc. Chem. Res.
2011, 44, 1329 – 1338.
[6] a) J. Rinnenthal, J. Buck, J. Ferner, A. Wacker, B. Fꢂrtig, H.
Schwalbe, Acc. Chem. Res. 2011, 44, 1292 – 1301; b) L. Chen, E.
Cressina, F. J. Leeper, A. G. Smith, C. Abell, ACS Chem. Biol.
2010, 5, 355 – 358.
ate exchange regime, and the folding of the aptamer into
a nearly bound conformation (Figure 5B). These experiments
clearly show that a reporter with high affinity for the RNA
target should be avoided for monitoring conformationally
labile RNA such as riboswitches, since it can dramatically
alter their native conformational equilibrium.
Interestingly, the 19F NMR signal of compound 1 is
detectable upon binding to the neomycin aptamer (Fig-
ure 6A) and shows signal splitting owing to the binding of
each compound 1 enantiomer to the aptamer.[8] A complete
19F NMR signal coalescence to the 19F signal of the free
compound 1 was observed upon addition of neomycin to the
mixture, which shows that the fluorinated probe was fully
displaced (Figure 6B). Moreover, 1H NMR spectroscopy
revealed a profound change in the aptamer structure that
now shows a neomycin-bound conformation (Figure 6E).
These experiments demonstrate that the competitive dis-
placement of a small fluorinated binder can be used to easily
monitor the conformational capture of a bi-stable RNA by its
ligand.
[7] C. Kreutz, H. Kꢃhlig, R. Konrat, R. Micura, Angew. Chem. 2006,
118, 3528 – 3532; Angew. Chem. Int. Ed. 2006, 45, 3450 – 3453.
[8] R. Moumnꢀ, M. Pasco, E. Prost, T. Lecourt, L. Micouin, C. Tisnꢀ,
J. Am. Chem. Soc. 2010, 132, 13111 – 13113.
[9] C. Dalvit, Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51, 243 –
271.
In conclusion, we have shown that small fluorinated
reporters are useful for monitoring the binding of unlabeled
RNA with unlabeled ligands. The observation of 19F NMR
chemical-shift variations can be used to qualitatively rank
different binders for a structured RNA, such as the 16S23
[10] a) C.-H. Wong, M. Hendrix, E. S. Priestley, W. A. Greenberg,
Chem. Biol. 1998, 5, 397; b) M. Kaul, C. M. Barbieri, D. S. Pilch,
J. Mol. Biol. 2005, 346, 119 – 134.
[11] The corresponding non-fluorinated diaminocyclopentanol is
a ligand of 16S23 RNA: L. Micouin, F. P. Dardel, C. Tisne-
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Vicrobeck, F. Maurice, M. Bonin, A. Perez Luna, C. J. Bour-
naud, G. Bꢀgis, WO2006024784, 2006.
[12] S. Yoshizawa, D. Fourmy, R. G. Eason, J. D. Puglishi, Biochem-
istry 2002, 41, 6263 – 6270.
[17] The KD measurements of high affinity binders (neomycin,
paromomycin) using imino proton chemical shifts were not
possible because of a slow exchange regime, even at 300 mm
KCl.
[13] The binding sites were localized using NOESYexperiments. See
Supporting Information for details.
[18] J. E. Weigand, M. Sanchez, E. B. Gunnesch, S. Zeiher, R.
Schroeder, B. Suess, RNA 2008, 14, 89 – 97.
[14] X. Zhang, A. Sꢃnger, R. Hemmig, W. Jahnke, Angew. Chem.
2009, 121, 6819; Angew. Chem. Int. Ed. 2009, 48, 6691.
[15] See Supporting Information.
[16] An alternative solution with probes undergoing intermediate
exchange can be used for the monitoring of the recovery of the
signal intensity upon adding of a competing molecule, as recently
proposed: A. Vulpetti, U. Hommel, G. Landrum, R. Lewis, C.
Dalvit, J. Am. Chem. Soc. 2009, 131, 12949 – 12959.
[19] E. Duchardt-Ferner, J. E. Weigand, O. Ohlenschlꢃger, S. R.
Schmidtke, B. Suess, J. Wçhnert, Angew. Chem. 2010, 122, 6352 –
6355; Angew. Chem. Int. Ed. 2010, 49, 6216 – 6219.
[20] All the 19F NMR spectroscopy experiments have been con-
ducted on a 300 MHz (H value) spectrometer with a standard
QNP probe. As 1 has been shown to bind to several structurally
different RNAs, we believe that it should be an excellent tool to
investigate many other RNAs.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
RNA Structures
T. Lombꢁs, R. Moumnꢂ, V. Larue, E. Prost,
M. Catala, T. Lecourt, F. Dardel,
L. Micouin,* C. Tisnꢂ*
&&&& — &&&&
Investigation of RNA–Ligand Interactions
by 19F NMR Spectroscopy Using
Fluorinated Probes
Spy swap: The interaction between an
unlabeled RNA and unlabeled ligands
(red hexagon) can be monitored by
19F NMR spectroscopy using small fluo-
rinated diamines (green star) as spy
reporters (see scheme). This technique
also enables the visualization of the
conformational capture of a riboswitch by
its ligand.
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!