Synthesis of spin-labeled riboswitch RNAs using convertible nucleosides and DNA-catalyzed RNA ligation

Article abstract of DOI:10.1016/j.bmc.2013.04.007

Chemically stable nitroxide radicals that can be monitored by electron paramagnetic resonance (EPR) spectroscopy can provide information on structural and dynamic properties of functional RNA such as riboswitches. The convertible nucleoside approach is used to install 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO) and 2,2,5,5-tetramethylpyrrolidin-1-oxyl (proxyl) labels at the exocyclic N⁴-amino group of cytidine and 2′-O-methylcytidine nucleotides in RNA. To obtain site-specifically labeled long riboswitch RNAs beyond the limit of solid-phase synthesis, we report the ligation of spin-labeled RNA using an in vitro selected deoxyribozyme as catalyst, and demonstrate the synthesis of TEMPO-labeled 53 nt SAM-III and 118 nt SAM-I riboswitch domains (SAM = S-adenosylmethionine).

Full text of DOI:10.1016/j.bmc.2013.04.007

Bioorganic & Medicinal Chemistry 21 (2013) 6171–6180
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis of spin-labeled riboswitch RNAs using convertible
nucleosides and DNA-catalyzed RNA ligation
⇑
Lea Büttner , Jan Seikowski , Katarzyna Wawrzyniak, Anne Ochmann, Claudia Höbartner
Research Group Nucleic Acid Chemistry, Max Planck Institute for Biophysical Chemistry, Am Fassberg 11, Göttingen 37077, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 15 April 2013
Chemically stable nitroxide radicals that can be monitored by electron paramagnetic resonance (EPR)
spectroscopy can provide information on structural and dynamic properties of functional RNA such as
riboswitches. The convertible nucleoside approach is used to install 2,2,6,6-tetramethylpiperidin-1-oxyl
(TEMPO) and 2,2,5,5-tetramethylpyrrolidin-1-oxyl (proxyl) labels at the exocyclic N4-amino group of
cytidine and 20-O-methylcytidine nucleotides in RNA. To obtain site-specifically labeled long riboswitch
RNAs beyond the limit of solid-phase synthesis, we report the ligation of spin-labeled RNA using an
in vitro selected deoxyribozyme as catalyst, and demonstrate the synthesis of TEMPO-labeled 53 nt
SAM-III and 118 nt SAM-I riboswitch domains (SAM = S-adenosylmethionine).
Keywords:
Solid-phase synthesis
Convertible nucleoside
Spin label
RNA ligation
Deoxyribozyme
Riboswitch
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
metabolite concentrations.10,11 The lengths of riboswitch domains
that are used for biophysical studies aiming at understanding ribo-
The structural and functional diversity of RNA is impressively
represented by the large number of different noncoding RNAs, that
have been discovered and continue to be discovered in all forms of
life.1 Spectroscopic studies of RNA folding and interaction of RNA
with proteins, metabolites and other small molecules play funda-
mental roles for elucidating mechanistic principles of RNA func-
tion. Biophysical techniques that are based on fluorescence or
EPR spectroscopy require the site-specific installation of emissive
or paramagnetic labels into the RNA of interest. A number of
synthetic approaches are available to directly incorporate modifi-
cations or install desired probes post-synthetically on pre-func-
tionalized RNA.2 Prominent methods involve conjugation of
fluorophores or spin labels via amide bond formation using acti-
vated esters and amino-modified nucleic acids, and recently also
cycloaddition reactions between alkynes and azides (click chemis-
try).3,4 Spin labels have also been incorporated into RNA via Pd-cat-
alyzed cross-coupling with halogenated nucleobases5 and via
conjugation to phosphorothioate-modified RNA.6 Convertible
nucleosides7 provide a well-established alternative to functionali-
zation of RNA at the exocyclic amino groups of the nucleobases.8
Recently, the first example for synthesis of spin-labeled RNA using
a nitroxide-modified nucleoside phosphoramidite was reported.9
A heavily investigated class of regulatory RNA elements are
riboswitch RNAs that alter gene expression levels in response to
switch mechanisms range from ꢀ30 to ꢀ150 nt,12 and are mostly
beyond the length that is easily achieved by solid-phase synthesis
in one piece. Therefore, synthetic modified RNA fragments are
enzymatically ligated to the desired full-lengths products.13 T4
DNA ligase and T4 RNA ligase have been successfully used to syn-
thesize chemically modified riboswitch RNAs14–18 and numerous
other large RNA targets, including for example ribozyme do-
mains19–21 and snRNAs.22 To our knowledge, the ligation of nitrox-
ide-labeled RNAs has not yet been reported. A potential reason
may be the sensitivity of nitroxides to reduction, a known problem
in EPR studies involving proteins,23 especially when the most
sensitive six-membered-ring nitroxide, 2,2,6,6-tetramethylpiperi-
din-1-oxyl (TEMPO) is employed. Nitroxide labels based on
five-membered heterocycles, such as 2,2,5,5-tetramethylpyrroli-
din-1-oxyl (proxyl) derivatives, are known to be more stable
towards reduction.24 Increasing the size of the geminal alkyl
groups (ethyl instead of methyl) has been shown to increase the
half-live of paramagnetic species in the presence of reductants.25,26
The stability of nitroxides has important implications for EPR
spectroscopy of nucleic acids in vivo, which is an emerging field
of research.27,28 With these and other recent advancements for
EPR-spectroscopic investigations of biomolecules,29,30 the accessi-
bility of long spin-labeled RNA is of general interest.
We report the synthesis of nitroxide-labeled RNA, carrying
2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO) or 2,2,5,5-tetrameth-
ylpyrrolidin-1-oxyl (proxyl) spin labels at the exocyclic amino
group of cytidine and 20-O-methylcytidine, resulting in four types
of spin-labeled RNA (Fig. 1). Here, we expand our previous report8
⇑
Corresponding author. Tel.: +49 551 201 1685; fax: +49 551 201 1680.
E-mail address: claudia.hoebartner@mpibpc.mpg.de (C. Höbartner).
These authors contributed equally to this work.
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.04.007

6172
L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
catalyst for the ligation of various spin-labeled and unmodified
RNA substrates. We demonstrate five different systems, using
RNA substrates in the lengths of 9–100 nt, resulting in ligation
products of 26–118 nt. In particular, we synthesized spin-labeled
variants of two riboswitch RNAs, SAM-I and SAM-III that bind the
universal cofactor S-adenosylmethionine (SAM).
2. Results and discussion
2.1. Synthesis of convertible uridine phosphoramidites
Convertible nucleosides allow the functionalization of nucleo-
bases at the exocyclic amino groups by displacement of appropri-
ate leaving groups with alkylamines. The incorporation of
O4-(4-chlorophenyl)uridine into RNA yields N4-modified cytidines
in the final RNA.7 The 20-O-tBDMS-protected phosphoramidite of
O4-(4-chlorophenyl)uridine is commercially available, and has
been employed successfully for the installation of TEMPO groups
in RNA.8 Here we report an analogous convertible uridine building
block that contains the (triisopropylsilyl)oxymethyl (TOM) group
as alternative fluoride-labile 20-protecting group. This permanent
RNA protecting group offers advantages compared to tBDMS in
terms of avoiding any issues related to potential 20,30-isomeriza-
tion and provides high coupling yields in shorter coupling times36
(generally we use 4 min for TOM-protected phosphoramidites and
12 min for tBDMS-protected phosphoramidites).
Figure 1. Convertible (4-chlorophenyl)uridine, TEMPO- and proxyl-labeled cyti-
dine/20-O-methylcytidine nucleotides in RNA. R = CH3 or TOM, R0 = 2-cyanoethyl.
on the installation of TEMPO groups on RNA using convertible
nucleosides in solid-phase synthesis. In addition, we report O4-
(4-chlorophenyl)-20-O-methyluridine phosphoramidite that yields
spin-labeled 20-O-methyl-cytidine in RNA. The 20-O-methyl group
is a well-known RNA modification that is of natural importance
and additionally has practical advantages due to the increased sta-
bility of modified RNAs.31 We examined and compared TEMPO-
and proxyl-labeled RNAs by analysis of UV melting curves, and
we characterized single-strand and duplex conformations by CW-
EPR spectroscopy. Furthermore, we demonstrate efficient ligation
of spin-labeled RNA by an optimized deoxyribozyme, as an alterna-
tive to protein-based RNA ligation.
Deoxyribozymes are artificial single-stranded DNAs of practical
utility that are the result of in vitro selection experiments.32,33 The
9DB1 deoxyriboyzme34 catalyzes the formation of a native 30–50
phosphodiester bond between two RNA fragments, by activating
the 30-hydroxyl group of the acceptor fragment for the nucleophilic
attack at the alpha phosphate of a 50-triphosphorylated donor frag-
ment.34 Recently we have identified the essential nucleotides in
the catalytic core of the original 9DB1 DNA by combinatorial muta-
tion interference analysis, and minimized the number of necessary
nucleotides.35 The shortened DNA 9DB1⁄ is here used as a general
The 20-O-TOM-protected phosphoramidite 4 was synthesized in
four steps from uridine, taking advantage of a late derivatization of
the nucleobase (Scheme 1). The 50-O-(4,40-dimethoxytrityl) (DMT)
group and the 20-O-TOM groups were installed following reported
procedures to yield nucleoside 2.36 The nucleobase was activated
by formation of the O4-(2,4,6-triisopropylphenyl)sulfonyl (=trisyl)
ester, which was not purified, and directly used in the next step.
In the presence of N,N-dimethylethylamine and DBU, the sulfonyl
ester was displaced with 4-chlorophenol to yield compound 3 in
excellent yield. Intermediate protection of the free 30-OH group
was not necessary; the presence of bulky DMT and TOM groups
and the large size of the trisyl group prevent any undesired side
Scheme 1. Synthesis of convertible uridine phosphoramidites 4 and 9. Reagents and condition. (i), DMT-Cl, pyridine, 6 h, rt; (ii), (a) nBu2SnCl2, iPr2NEt, (CH2Cl)2, 45 min, rt, (b)
TOM-Cl, 20 min, 80 °C.; iii, (a) trisyl-Cl, DMAP, NEt3, CH2Cl2, 2 h, 0 °C–rt; (b) 4-chlorophenol, Me2NEt, DBU, 3 h, rt; (iv) CEP-Cl, Me2NEt, CH2Cl2, 2.5 h, rt; (v), Ph2CO3, NaHCO3,
DMF, 4 h, 120 °C; (vi), Mg(OCH3)2, DMF, 5 h, 100 °C.

L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
6173
reaction during sulfonylation. The 30-OH was then functionalized
using 2-cyanoethyl-(N,N-diisopropylamino) chlorophosphite
(CEP-Cl) and N,N-dimethylethylamine in dichloromethane to give
phosphoramidite 4. Synthesis of the O4-(4-chlorophenyl)-20-O-
methyluridine phosphoramidite 9 was achieved along a similar
route. Starting from uridine, 2,20-O-anhydrouridine (5) was gener-
ated and converted to the 50-O-DMT-protected compound 6 fol-
lowing reported procedures.37,38 Reaction with magnesium
methoxide yielded nucleoside 7,39,40 which was subjected to O4
activation as trisyl ester and derivatization with 4-chlorophenol.
Without intermediate protection of the 30-OH, the new compound
8 was obtained in respectable yield, and was then converted to 2-
cyanoethyl N,N-diisopropyl phosphoramidite 9 upon reaction with
CEP-Cl under established conditions.
2.2. Synthesis of spin-labeled RNA
The phosphoramidites 4 and 9 were used for solid-phase syn-
thesis of RNA oligonucleotides on polystyrene support using 20-
O-TOM-protected RNA phosphoramidites, employing standard
RNA coupling conditions and using S-benzylthiotetrazole (BTT) as
activator. After complete assembly of the desired oligonucleotide
sequence, the solid support was incubated with a 2 M solution of
either 4-amino-TEMPO or 3-amino-proxyl in methanol at 42–
55 °C for 24–48 h, followed by a short incubation with MeNH2 or
NH4OH in H2O for 2–5 h to remove remaining base-labile protect-
ing groups (this step can be skipped for short TEMPO-labeled oligo-
ribonucleotides with less than 15 nt, in which case the excess of
amine is sufficient to cleave acyl and cyanoethyl groups). The 20-
protecting groups were removed with 1 M TBAF in THF, followed
by desalting of the crude product and purification of the desired
RNA by PAGE or anion exchange HPLC. The sequences of the syn-
thesized spin-labeled oligonucleotides ranging from 9 to 27 nt
are summarized in Table 1. In this manuscript, unmodified oligo-
nucleotides are assigned with bold numbers. Modifications are
indicated by lower case letters following the oligonucleotide num-
ber: a = CTEMPO, b = Cproxyl, c = CmTEMPO, d = Cmproxyl (Fig. 1). In
addition to the nitroxide reagents, we have used 4-amino-
2,2,6,6-tetramethylpiperidine (pip, letter e in oligo names) for
substitution of O4-chlorophenyl uridine to prepare reference com-
pounds for the stably reduced state of the TEMPO label (19e and
22e).
Figure 2. Synthesis, purification and analysis of spin-labeled RNA. (a) TEMPO-
labeled 15-mer 11c; (b) proxyl-labeled 15-mer 11d. Anion exchange HPLC of crude
and purified oligonucleotides, Dionex DNA-Pac PA200, 4 ꢁ 250 mm, 70 °C, 6 M urea,
0–48% in 12 CV. Denaturing polyacrylamide gel (7 M urea), 0.7 ꢁ 200 ꢁ 320 mm,
35 W. ESI-MS: raw spectrum (left) and deconvoluted mass spectrum (right,
excerpt).
yield. In contrast, substitution with 3-amino-proxyl was generally
less efficient; two products of about equal intensity were pro-
duced, which could be easily separated by denaturing polyacryl-
amide gel electrophoresis (see inset in Fig. 2b). The lower band
(earlier eluting product on anion exchange column) was identified
by ESI-MS as the N4-methylated derivative (m4Cm, m4C respec-
tively), which originates from substitution of the O4-chlorophenyl
group with MeNH2. All spin-labeled RNAs were characterized by
ESI-MS and the purity was checked by analytical anion-exchange
HPLC (see supporting Figures S1-S3 for more examples).
Figure 2 displays examples for anion exchange HPLC profiles of
crude and purified 15-mer RNAs 11c and 11d containing TEMPO-
and proxyl-modified 20-O-methylcytidine. The substitution with
4-amino-TEMPO was highly efficient and the desired TEMPO-la-
beled RNAs were produced as the major products in very good
2.3. Characterization of spin-labeled RNA by UV melting
analysis and EPR spectroscopy
The influence of the TEMPO and proxyl spin labels on the ther-
modynamic stability of RNA duplexes was studied by analysis of
UV melting curves. Three different lengths of RNA duplexes were
investigated, containing 9, 15, or 20 base-pairs. The thermal melt-
ing results are summarized in Table 2, and melting curves for
unmodified, TEMPO- and proxyl-labeled duplexes 11/110 are de-
picted in Figure 3 and Figure S4. Thermodynamic parameters were
extracted for the 15-mer duplexes 11/110 by analyzing melting
experiments at seven different concentrations (see Tables S3, S4,
and Figure S5). All other duplexes were studied at least at two dif-
ferent RNA concentrations. The influence of the TEMPO and proxyl
spin labels on the duplex stability is comparable; the TEMPO label
is only minimally favored by a small difference of less than 1 °C
(compare samples a with b and c with d in Table 2). The 20-O-
methyl group in samples c/d shows the expected small stabilizing
effect compared to a/b for the 15 and 20 bp duplexes, but the influ-
ence on the 9 bp duplex is negligible. The drop in melting temper-
ature compared to the unmodified RNA depends on the length of
the duplex for all four modifications, and amounts to 8–9 °C for
Table 1
Modified RNA oligonucleotides

6174
L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
Table 2
UV-melting analysis of spin-labeled RNA duplexes
aDTm = Tm(spin-labeled RNA)—Tm(unmodified RNA); in 10 mM potassium phosphate buffer, pH 7.0, 150 mM NaCl.
Figure 4. CW-EPR characterization of (a) TEMPO- and (b) proxyl-labeled RNAs 11
in single-strand (black) and duplex (red) in 2.5 mM potassium phosphate buffer, pH
7.0, 30 mM NaCl.
Figure 3. UV melting curves of spin-labeled RNA duplexes 11/110. 4
10 mM potassium phosphate buffer, pH 7.0, 150 mM NaCl. (a) Hyperchromicity at
250 nm, (b) derivative of hyp, maximum gives Tm
lM RNA, in
.
of a native 30–50 phosphodiester bond between two RNA substrates
for the ligation of spin-labeled oligonucleotides. We also compare
the efficiency of DNA-catalyzed ligation with the traditionally used
splinted ligation with T4 DNA ligase, and we study the fate of the
spin-labeled species under ligation conditions and in the purified
ligation product. Ligation experiments of spin-labeled RNAs are
summarized in Table 3, the sequences of unmodified acceptor
RNAs, 50-activated donor RNAs, corresponding deoxyribozymes
(D1–D5), and DNA splints are given in Tables S1 and S2.
the shortest 9 bp strands, and 5–6 °C for the samples containing 15
and 20 bp.
CW-EPR spectra were recorded to characterize the spin-labeled
RNA oligonucleotides (Fig. 4 and Fig. S6), and to compare the rela-
tive mobility of the TEMPO and proxyl labels. The observed spec-
tral broadening in the presence of the complementary strand
confirms duplex formation. Compared to the non-base-paired situ-
ation in the single-strand, the overall dynamics of the spin label is
reduced upon formation of a Watson–Crick base-pair between the
labeled cytidine residue and guanosine in the complementary
strand. The broader spectra for 11d compared to 11c suggest that
the proxyl label is less mobile than the TEMPO label (see Fig. S6).
The 9DB1⁄ deoxyribozyme (Fig. 5a) hybridizes the RNA sub-
strates via the 50 and 30 binding arms and uses the nucleotides of
the catalytic core together with divalent metal ion cofactors to
activate the 30-OH group of the acceptor RNA for nucleophilic at-
tack on the 50-triphosphate of the donor RNA. Upon release of
pyrophosphate, the native phosphodiester linkage is formed. The
ligation sites for DNA-catalyzed RNA ligation were chosen based
on the reported sequence preferences of 9DB1: the 30-terminus
of the acceptor RNA can be any nucleotide except C, and transcrip-
tion of the donor RNA should be initiated with 50-GA or 50AA (i.e.,
the sequence requirement is D|RA).34
2.4. Ligation of short spin-labeled RNA
Using EPR to study longer RNAs than can be achieved by solid-
phase synthesis requires ligation of modified RNA fragments. Here
we employ the deoxyribozyme 9DB1⁄ that catalyzes the formation
Table 3
Overview of RNA substrates and conditions for DNA-catalyzed ligation of spin-labeled RNA (spin-labeled nucleotides marked in red)
aCondition A: 40 mM Mg2+, pH 9.0, 37 °C, 5 h; B: 20 mM Mn2+, pH 7.5, 37 °C, 5 h; C: 20 mM Mn2+, pH 7.5, 25 °C, 15 h.

L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
6175
Figure 5. (a) Schematic depiction of RNA ligation by 9DB1⁄ deoxyribozyme. (b) Analysis of ligation kinetics with Mn2+ as cofactor. (c) Kinetics plot (single-turnover
conditions); kobs(Mn2+) = 0.15 minꢂ1, kobs(Mg2+) = 0.04 minꢂ1. (d) PAGE separation of ligation reactions 11+15 with 9DB1⁄ DNA enzyme D2 (1.5 nmol scale). Lane 1. reference;
lane 2. unmodified acceptor 11; lane 3: CmTEMPO-labeled acceptor 11c, lane 4: Cmproxyl-labeled acceptor RNA 11d. (e) Anion exchange HPLC analysis of ligation reaction after
5 h incubation time, and HPLC trace of gel-purified Cmproxyl-labeled ligation product 16d (isolated yield: 38%). (f) ESI-MS of 16d.
The minimized deoxyribozyme 9DB1⁄ used in this study re-
sulted from combinatorial mutation interference analysis of the
original sequence.35 Upon continued studies on the mechanism
of the DNA-catalyzed reaction, we found that Mn2+ at pH 7.5 leads
to faster ligation compared to the originally reported conditions
with Mg2+ at pH 9.0 (Fig. 5). For working with longer RNA sub-
strates, the lower pH is an additional advantage. The ligations were
scaled up and performed in parallel with unmodified and spin-la-
beled RNAs. In all studied cases involving RNAs 10, 11 and 12,
the spin labels did not affect the reaction, and the paramagnetic
products were isolated by denaturing PAGE, and characterized by
HPLC, MS and CW-EPR.
ysis of spin-labeled RNA after incubation with T4 DNA ligase buffer
in contrast to deoxyribozyme ligation buffer (Fig. 7a). While the
RNA was still intact after 5 h of incubation in the presence of
20 mM Mn2+ at pH 7.5, about 15% of the corresponding amine
was produced in the presence of 10 mM DTT in the T4 DNA ligase
buffer. In a second experiment, TEMPO-labeled RNA was incubated
with 100 mM DTT in 10 mM Tris–HCl, pH 7.8 and analyzed by an-
ion exchange HPLC (Fig. 7b). By comparison and co-injection
(Fig. S9) with an authentic sample of 2,2,6,6-tetramethylpiperi-
dine-substituted RNA it was confirmed that the undesired byprod-
uct must have lost the nitroxide. T4 DNA ligase can also be used in
a DTT-free reaction buffer, but the concentration of DTT that comes
from the storage buffer of the enzyme was sufficient to complicate
the purification of high-quality spin-labeled RNA. The instability of
TEMPO groups in the presence of DTT is a known problem that has
affected investigations of spin-labeled proteins.23 The effect of DTT
concentration on the intensity of the EPR signal has been reported,
and tris(2-carboxyethyl)phosphine (TCEP) has been suggested as
alternative sulfhydryl reductant that leads to reduced rate of TEM-
PO reduction.23 Potentially TCEP could be applied in T4 DNA cata-
lyzed reactions. However, the general applicability of the 9DB1⁄
deoxyribozyme, as demonstrated here for several examples of
spin-labeled RNAs of different lengths, offers an efficient and
reductant-free alternative for enzymatic ligation of labeled
RNAs.
Before studying 9DB1⁄ for the ligation of spin-labeled RNA, we
used traditional splinted ligation with T4 DNA ligase (or T4 RNA li-
gase) for connecting synthetic RNA fragments. However, using
spin-labeled substrates, we faced the formation of a number of
undesired byproducts which caused a reduced yield of the desired
products. In addition, upon gel-purification, a small amount of a
byproduct that elutes earlier on anion exchange HPLC could not
be separated. This byproduct contained a reduced TEMPO substitu-
ent. In contrast, the ligation reaction with 9DB1⁄ proceeded very
cleanly and produced spin-labeled ligation product in high quality
and good yield (Fig. 6, S7, S8).
To further investigate the problematic partial loss of spin label
during protein-catalyzed ligation reactions, we compared the anal-

6176
L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
Figure 7. Examination of stability of spin-labeled RNA under ligation conditions. (a)
20-mer CTEMPO-labeled RNA 12a is incubated with T4 DNA ligase buffer (40 mM
Tris–HCl pH 7.8, 10 mM MgCl2, 10 mM DTT, 0.5 mM ATP) or with deoxyribozyme
buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 2 mM KCl, 20 mM MnCl2) for 5 h at
37 °C. (b) Characterization of reduction product by comparison with authentic
2,2,6,6-tetramethylpiperidine-substituted RNA 19e.
Figure 6. Comparison of protein- and DNA-catalyzed RNA ligation. (a) T4-DNA
ligase mediated splinted ligation of CTEMPO-labeled RNA 12a and 50-phosphorylated
donor RNA 17. conditions: 15
10 mM DTT, 0.5 mM ATP, 37 °C, 12 h. (b) 9DB1⁄-catalyzed ligation of 12a and 50-
triphosphorylated donor RNA 17. Conditions: 15 M RNA, 20 mM MnCl2, pH 7.5,
150 mM NaCl, 2 mM KCl, 37 °C, 5 h. The product produced by T4 DNA ligase could
not be purified completely by PAGE (<10% reduced product present(⁄)). The isolated
yield of ligation product 18a was 24% for (a) and 42% for (b).
lM RNA, 40 mM Tris–HCl, pH 7.8, 10 mM MgCl2,
l
tion site resides within helix P2. Two 50-triphosphorylated donor
substrates were generated by in vitro transcription. RNA 23 spans
nt 19–70 of the riboswitch RNA and was used as a test case to con-
firm the efficiency of the ligation site (Fig. S10). The 100 nt donor
substrate 25 was generated by in vitro transcription from a dou-
ble-stranded DNA template that was produced by polymerase
extension of overlapping primers. The DNA-catalyzed ligation
was performed at pH 7.5 in 50 mM 4-(2-hydroxyethyl)-1-piperaz-
ineethanesulfonic acid (HEPES) buffer in the presence of 20 mM
2.5. Synthesis of spin-labeled SAM riboswitch RNAs
To further demonstrate the general applicability of the DNA-
catalyzed ligation of spin-labeled RNA, we focused on the prepara-
tion of TEMPO-labeled riboswitch RNAs that may enable studies of
RNA dynamics upon ligand binding by EPR spectroscopy. The first
target is the 53 nt SAM-III/SMK riboswitch, an analog of the SMK se-
quence from Enterococcus faecalis that was optimized for determi-
nation of the crystal structure (Fig. 8a).41 We decided to split the
riboswitch into two pieces of similar size and chose a ligation site
within the non-conserved stem-loop region P3. In the model sys-
tems of the previous section, the spin label was installed distant
from the ligation site. To employ pulsed EPR methods, it will be
necessary to install two spin labels in the target RNA. Therefore
we also explored a labeling position that is closer to the ligation
site. In the SAM-III RNA, we chose cytidine C4 in P1 and C23 in
P3 for labeling. The two 27 nt spin-labeled RNAs 20a1 and 20a2
were prepared as described above and purified by PAGE. The sec-
ond fragment of 26 nt (donor substrate) was prepared by in vitro
transcription. The 9DB1⁄-catalyzed ligation reaction was per-
formed with equimolar concentrations of both substrates and
deoxyribozyme, in 50 mM N-cyclohexyl-2-aminoethansulfonic
acid (CHES) buffer, pH 9.0, containing 150 mM NaCl, 2 mM KCl,
and 40 mM MgCl2, at 37 °C for 5 h. After addition of EDTA and pre-
cipitation, the nucleic acids were separated by PAGE (Fig. 8b). The
ligation efficiency was similar in all three reactions, indicating that
the spin labels were well tolerated at both investigated positions.
The ligated RNA was isolated and characterized by CW-EPR in com-
parison to the non-ligated acceptor RNA 20a2 (Fig. 8c). The slightly
broader spectrum of the full-length 53-mer RNA suggests reduced
motion compared to the shorter unstructured 27-mer, as it would
be expected upon almost doubling the size of the RNA.
MnCl2. Examination of the ligation time-course at 10 lM RNA con-
centration revealed good conversion and produced 70% ligated
product after 3–5 h. Figure 8e displays the analysis of the ligation
reaction by comparing the sample composition at the start of the
ligation and after 5 h of incubation at 37 °C for the unmodified
(22) and the TEMPO-labeled substrate 22a. In a separate larger
scale reaction, the full-length spin-labeled SAM-I riboswitch RNA
was isolated in 30% yield (Fig. S11). These results pave the way
for EPR-based studies of changes in local dynamics upon binding
of ligands to long riboswitch RNAs.44,45
3. Conclusions
In summary, we have introduced TEMPO and proxyl labels at N4
of cytidine and 20-O-methylcytidine, and used the resulting spin-la-
beled RNA oligonucleotides as acceptor substrates in DNA-catalyzed
ligation reactions for the synthesis of long spin-labeled RNA. Both
TEMPO and proxyl labels are well accommodated in the major
groove of RNA duplexes, as revealed by only moderate thermody-
namic destabilization and the observed spectral broadening of
CW-EPR signals upon hybridization to complementary strands.
Spin-labeled RNAs are suboptimal substrates for ligation by T4
DNA ligase due to the sensitivity of nitroxides to reduction. Impor-
tantly, under the conditions used for DNA-catalyzed ligation with
9DB1⁄, the spin label integrity was not affected. The TEMPO and
proxyl groups were well tolerated at diverse positions in the accep-
tor strand that is generated by solid-phase synthesis using convert-
ible nucleoside phosphoramidites. The donor strand is accessible
by in vitro transcription, and is currently not directly amenable
to site-specific modification using convertible nucleosides. How-
As a second target, we chose the 118-mer yitJ SAM-I riboswitch
RNA of Bacillus subtilis (Fig. 8d).42,43 The TEMPO label was intro-
duced at position 8 of the 18-nt 50-terminal fragment 22a. The liga-

L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
6177
Figure 8. Synthesis of spin-labeled riboswitch RNAs. (a) fragment of SAM-III/SMK riboswitch, analogous to construct used for determination of X-ray structure.41 Blue spheres
in the pymol picture indicate attachment points for TEMPO (i.e., N4 of C4 and C23), (b) gel image of SAM-III ligation: lane 1. reference for donor 21; lane 2. ligation of
unmodified RNAs 20 + 21, lane 3. TEMPO at C4: 20a1 + 21; lane 4. TEMPO at C23: 20a2+21; conditions for lanes 2–4: 15
deoxyribozyme D4. (c) CW-EPR spectrum of 53 nt SAM-III RNA (with TEMPO at C23) in comparison to unligated RNA 20a2 (40
of Bacillus subtilis yitJ SAM-I riboswitch aptamer (118 nt);43 (d) gel image of SAM-I ligation: lane 1. reference for donor substrate 25 (100 nt); ligation of spin-labeled SAM-I
l
M RNA, 40 mM MgCl2, pH 9.0, 37 °C, 5 h; 9DB1⁄
M RNA in each case). (d) Schematic depiction
l
22a + 25: lane 2: 5 h. lane 3: start; unmodified comparison 22 + 25: lane 4: start, lane 5: 5 h; lane 6: reference for acceptor 22; ligation conditions lanes 2–5: 15
lM RNA,
20 mM MnCl2, pH 7.5, 37 °C, 5 h; 9DB1⁄ deoxyribozyme D5.
ever, with recent reports of improved synthetic methods for the
installation of 50-triphosphates by solid-phase synthesis,46,47 func-
tionalized and/or labeled donor substrates for 9DB1⁄ ligation are
coming into reach.
obtained by the described method; ligation of spin-labeled RNA
was here performed on the low nmol-scale (up to 3 nmol), but
the approach is expected to be easily scalable to multi-nmols of
isolated ligation product. Moreover, the application of 9DB1⁄ on
preparative scale is certainly not limited to TEMPO- and proxyl-la-
beled RNA, but will be useful for the synthesis of diverse labeled
RNAs for spectroscopic and biochemical studies.
With the successful DNA-catalyzed synthesis of spin-labeled
RNAs up to 118 nt in length, this work sets a new benchmark for
preparative applications of deoxyribozymes. The DNA catalyst for
a ligation reaction can be obtained as standard DNA oligonucleo-
tide from commercial suppliers, just like DNA splints or primers
are nowadays ordered by scientists on an almost daily basis. Sim-
ple buffers and divalent metal ions are the only other necessary
ingredients. The DNA catalyst is used in stoichiometric amount
with respect to the RNA substrates, and can easily be separated
from the product, re-isolated and used again in another ligation
experiment.
4. Experimental procedures
4.1. General procedures
Thin layer chromatography (TLC) was carried out using silica
gel on aluminum sheets (200 lM, F-254). Compounds were visu-
alized by UV light and staining with p-anisaldehyde. Flash-col-
umn chromatography was performed using ultrapure flash
silica gel (230–400 mesh size, 60 Å). All moisture- and air-sensi-
tive reactions were carried out in oven-dried glassware under an
One of the advantages of EPR spectroscopy lies in the relatively
small amount of sample that is required. With modern instruments
and high-field techniques, EPR experiments can be performed with
less than 5 lL of 50 l
M spin-labeled RNA.48 This amount is readily
inert argon atmosphere. NMR spectra were recorded on
a

6178
L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
400 MHz spectrometer. Commercial grade CDCl3 was passed over
basic alumina shortly before use for tritylated compounds. 1H
NMR chemical shifts are reported in reference to undeuterated
residual solvent in CDCl3 (7.26 ppm). 13C NMR chemical shifts
are reported in reference to solvent signal (CDCl3 (77.16 ppm).
31P NMR chemical shifts are reported relative to 85% H3PO4 as
an external standard.
rated under reduced pressure. The crude product was purified by
column chromatography on SiO2 (hexane/ethyl acetate 3:1, 2%
NEt3) yielding compound 4 (159 mg, 0.15 mmol, 64%) as a white
foam. Rf (hexane/ethyl acetate 3:1) = 0.39; 1H NMR (400 MHz,
CDCl3): d (ppm) = 1.00 (d, 12 H, J = 6.7 Hz, (CH3)2CHN), 1.03–1.19
(m, 54 H, (CH3)2CHN, ((CH3)2CH)2Si), 2.40 (t, 2 H, J = 6.4 Hz,
OCH2CH2CN-Ha), 2.60 (t, 1 H, J = 6.2 Hz, OCH2CH2CN-Hb, diastereo-
mer 1), 2.61 (t, 1 H, J = 6.2 Hz, OCH2CH2CN-Hb, diastereomer 2),
3.46 (td, 2 H, J = 10.6 Hz, J = 2.5 Hz, H-C(50)a), 3.52–3.75 (m, 8 H,
H-C(50)b, 2 ꢁ (CH3)2CH-N, OCH2CH2CN-Ha),3.81 (s, 12 H, OMe),
3.88–3.96 (m, 2 H, OCH2CH2CN-Hb), 4.25–4.32 (m, 4 H, H-C(20),
H-C(40)), 4.46 (ddd, 1 H, J = 9.5 Hz, J = 8.0 Hz, J = 4.9 Hz, H-C(30), dia-
stereomer 1), 4.57 (ddd, 1 H, J = 9.5 Hz, J = 8.0 Hz, J = 4.9 Hz, H-
C(30), diastereomer 2), 5.13 (d, 1 H, J = 4.2 Hz, 20-OCH2-Ha, diaste-
reomer 1), 5.14 (d, 1 H, J = 4.2 Hz, 20-OCH2-Ha, diastereomer 2),
5.22 (d, 1 H, J = 4.2 Hz, 20-OCH2-Hb, diastereomer 1), 5.23 (d, 1 H,
J = 4.2 Hz, 20-OCH2-Hb, diastereomer 2), 5.56 (d, 1 H, J = 7.3 Hz, H-
C(5), diastereomer 1), 5.62 (d, 1 H, J = 7.3 Hz, H-C(5), diastereomer
2), 6.12(s, 2 H, H-C(10)), 6.83–6.88 (m, 8 H, DMT-Har), 7.06–7.09 (m,
4 H, pCl-Ph-Har), 7.25–7.34 (m, 18 H, Har), 7.39–7.45 (m, 4 H, Har),
8.42 (d, 1 H, J = 7.4 Hz, H-C(6), diastereomer 1), 8.49 (d, 1 H,
J = 7.4 Hz, H-C(6), diastereomer 2); 31P NMR (162 MHz, CDCl3): d
(ppm) = 149.9 (s); MS (ESI) m/z = 1065 [M+Na]+.
4.2. Synthesis of convertible nucleoside phosphoramidites
4.2.1. 50-O-(4,40-Dimethoxytrityl)-20-O-(triisopropylsilyl)oxymethyl-
O4-(4-chlorophenyl)uridine (3)
50-O-DMT-20-O-TOM-uridine 2 was prepared from uridine in
two steps as previously described.36 Compound
2 (390 mg,
0.53 mmol, 1.00 equiv) was dissolved in DCM (6 mL), and DMAP
(9.8 mg, 0.08 mmol, 0.15 equiv) and NEt3 (0.66 mL, 4.79 mmol,
9.00 equiv) were added. The solution was cooled to 0 °C and
2,4,6-triisopropylbenzenesulfonyl chloride (212 mg, 0.70 mmol,
1.32 equiv) was added. The solution was stirred at 0 °C for
10 min, allowed to warm up and stirred at ambient temperature
for 2 h. The reaction mixture was diluted with DCM, washed
with NaHCO3, dried over Na2SO4 and evaporated under reduced
pressure. The crude product was directly used for the next step.
Crude 50-O-DMT-20-O-TOM-O4-trisyl-uridine (532 mg) was dis-
solved in DCM (12 mL) and 4-chlorophenol (343 mg, 2.67 mmol,
5.00 equiv) and N,N-dimethylethylamine (0.87 mL, 8.00 mmol,
15.0 equiv) were added. DBU (0.08 mL, 0.53 mmol, 1.00 equiv)
was added over 5 min and the resulting mixture was stirred
for 90 min. The organic layer was washed with NaHCO3, dried
over Na2SO4 and evaporated under reduced pressure. Purification
by column chromatography on SiO2 with 20–80% ethyl acetate in
hexane (containing 2% NEt3) yielded the desired compound 3
(423 mg, 0.50 mmol, 94%) as white foam. Rf (hexane/ethyl ace-
tate 3:1) = 0.32; 1H NMR (400 MHz, CDCl3): d (ppm) = 1.04–1.09
(m, 21 H, (CH3)2CH-Si, (CH3)2CH-Si), 3.33 (d, 1 H, J = 8.4 Hz, HO-
C(30)), 3.56 (dd, 1 H, J = 11.2 Hz, J = 2.3 Hz, H-C(50)a), 3.62 (dd, 1
H, J = 11.2 Hz, J = 2.3 Hz, H-C(50)b), 3.81 (s, 6 H, OCH3), 4.08 (dt,
1 H, J = 8.9 Hz, J = 2.3 Hz, H-C(40)), 4.22 (d, 1 H, J = 5.0 Hz, H-
C(20)), 4.42 (ddd, 1 H, J = 8.9 Hz, J = 8.4 Hz, J = 5.0 Hz, H-C(30)),
5.12 (d, 1 H, J = 4.7 Hz, 20-OCH2O), 5.27 (d, 1 H, J = 4.7 Hz, 20-
OCH2O), 5.65 (d, 1 H, J = 7.4 Hz, H-C(5)), 5.94 (s, 1 H, H-C(10)),
6.86 (dd, 4 H, J = 8.6 Hz, J = 1.1 Hz, DMT-Har), 7.09 (dd, 2 H,
J = 9.6 Hz, J = 3.2 Hz, pCl-Ph-Har), 7.23–7.28 (m, 2 H, Har), 7.30–
7.35 (m, 7 H, Har), 7.40–7.44 (m, 2 H, Har), 8.50 (d, 1 H,
J = 7.4 Hz, H-C(6)); 13C NMR (100 MHz, CDCl3): d (ppm) = 11.8
((CH3)2CH-Si), 17.8 ((CH3)2CH-Si), 55.3 (DMT-OCH3), 61.0
(C(50)), 67.6 (C(30)), 83.3 (C(40)), 83.4 (C(20)), 87.0 (DMT-Cq),
90.1 (C(10)), 90.8 (20-OCH2O), 94.8 (C(5)), 113.3 (DMT-Car),
123.1 (pCl-Ph-Car), 123.8, 123.9, 127.1, 128.0, 128.3, 129.6,
130.3, 131.2, 135.2, 135.5, 144.5 (C(6)), 144.6, 150.1, 151.3,
155.0 (C(2)), 158.7 (DMT-Car), 158.7 (DMT-Car), 171.2 (C(4));
MS (ESI) m/z = 865.1 [M+Na]+.
4.2.3. 50-O-(4,40-Dimethoxytrityl)-20-O-methyluridine (7)
50-O-DMT-2,20-O-anhydrouridine (6) was prepared in two steps
from uridine as previously described.37,38 The conversion to 50-O-
DMT-20-O-methyluridine was performed in analogy to earlier re-
ports.39,40 Magnesium methoxide was freshly prepared by heating
Mg (0.5 g) in methanol (50 mL) at 60 °C for 2 h, evaporation of ex-
cess methanol and drying of the gray powder under vaccum. Dried
Mg(OCH3)2 (740 mg, 8.57 mmol, 6.00 equiv) was added to a sus-
pension of compound 6 (750 mg, 1.42 mmol, 1.00 equiv) in DMF
(20 mL) and the mixture was heated to 100 °C for 2 h. A clear solu-
tion was obtained. The solvent was evaporated and the residue was
taken up in ethyl acetate. The organic layer was washed with NaH-
CO3 and water, dried over Na2SO4 and evaporated under reduced
pressure. The crude product was purified by column chromatogra-
phy on SiO2 with 2–4% methanol in DCM to give compound 7
(510 mg, 0.91 mmol, 64%) as white solid. 1H NMR (400 MHz,
CDCl3): d (ppm) = 3.54–3.57 (m, 2 H, H-C(50)), 3.65 (s, 3 H, 20-
OCH3), 3.79-3.80 (m, 7 H, H-C(20), DMT-OCH3), 3.98–4.01 (m, 1 H,
H-C(40)), 4.45-4.50 (m, 1 H, H-C(30)), 5.26 (d, 1 H, J = 8.1 Hz, H-
C(5)), 5.97 (s, 1 H, H-C(10)), 6.82-6.86 (m, 4 H, DMT-Har), 7.24–
7.39 (m, 9 H, DMT-Har), 8.03 (d, 1 H, J = 8.1 Hz, H-C(6)); MS (ESI):
m/z = 583.1 [M+Na]+.
4.2.4. 50-O-(4,40-Dimethoxytrityl)-O4-(4-chlorophenyl)-20-O-
methyluridine (8)
50-O-DMT-20-O-methyluridine
(7)
(120 mg,
0.26 mmol,
1.00 equiv) was dissolved in DCM (1 mL), triethylamine (0.20 mL)
and DMAP (120 mg, 0.26 mmol, 1.00 equiv) were added at 0 °C
and the mixture was stirred at 0 °C for 15 min. 2,4,6-tri-
isopropylbenzenesulfonyl
chloride
(133 mg,
0.44 mmol,
1.70 equiv) was added and stirring was continued for 30 min at
0 °C and another 2 h at ambient temperature. The reaction mixture
was diluted with DCM and the organic layer washed with water
and NaHCO3, dried over Na2SO4 and evaporated under reduced
pressure. The residue was taken up in DCM (2 mL) and N,N-dim-
ethylethylamine (133 mg, 0.44 mmol, 1.70 equiv) and 4-chloro-
phenol (133 mg, 0.44 mmol, 1.70 equiv) were added under
vigorous stirring. After 20 min, DBU (39.6 mg, 0.26 mmol,
1.00 equiv) was added and stirring was continued for 3 h. The sol-
vent was evaporated under reduced pressure and the crude prod-
uct was purified by column chromatography on SiO2 with 2–5%
ethyl acetate in hexane to give target compound 8 (107 mg,
4.2.2. 50-O-(4,40-Dimethoxytrityl)-20-O-
(triisopropylsilyl)oxymethyl-O4-(4-chlorophenyl)uridine 30-(2-
cyanoethyl-N,N-diisoproyplphosphoramidite) (4)
50-O-DMT-20-O-TOM-O4-(4-chlorophenyl)uridine
3
(201 mg,
0.24 mmol, 1.00 equiv) was dried under high vacuum overnight
and dissolved in DCM (5 mL). N,N-Dimethylethylamine (0.26 mL,
0.24 mmol, 1.00 equiv) and 2-cyanoethyl-(N,N-diisopropylamino)
chlorophosphite (98.7 mg, 0.42 mmol, 1.75 equiv) were added
and the solution stirred at ambient temperature for 2.5 h. A few
drops of methanol were added to stop the reaction, and the organic
phase was washed with NaHCO3, dried over Na2SO4 and evapo-

L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
6179
0.16 mmol, 62%) as white solid. Rf (hexane/ethyl acetate
1:1) = 0.18; 1H NMR (400 MHz, CDCl3): d (ppm) = 2.56 (d, 1 H,
J = 10.5 Hz, HO-C(30)), 3.53–3.65 (m, 3 H, H-C(50), H-C(20)), 3.72 (s,
3 H, 20-OCH3), 3.81 (s, 6 H, OCH3), 3.99-4.04 (m, 1 H, H-C(40)),
4.42–4.51 (m, 1 H, H-C(30)), 5.64 (d, 1 H, J = 7.3 Hz, H-C(5)), 5.98
(s, 1 H, H C(10)), 6.84-6.88 (m, 4 H, DMT-Har), 7.06–7.10 (m, 2 H,
pCl-Ph-Har), 7.28–7.43 (m, 11 H, DMT-Har, pCl-Ph-Har), 8.54 (d,
vents were evaporated. The residue was dissolved in a solution of
TBAF in THF (500 L, 1 M) and incubated at room temperature for
8–16 h. Aqueous buffer (500 L, 1 M Tris–HCl pH 8.0) was added,
l
l
and THF was removed under reduced pressure before desalting
on HiTrap desalting columns (3 ꢁ 5 mL). The aqueous eluate was
concentrated and the crude product stored at ꢂ20 °C.
Anion exchange HPLC was performed on Dionex DNA-PAc
PA200 column, 4 ꢁ 250 mm at 70 °C. Buffer A: 25 mM Tris–HCl,
pH 8.0, 6 M urea. Buffer B: same as A + 0.5 M NaClO4.
1 H, J = 7.3 Hz, H-C(6)). 13C NMR (100 MHz, CDCl3):
d
(ppm) = 55.7 (DMT-OCH3), 59.3 (20-OCH3), 61.1 (C(50)), 68.1
(C(30)), 83.5 (C(20)), 84.2 (C(40)), 88.7 (Cq), 95.2 (C(10)), 113.5,
113.7 (DMT-Car), 123.5, 127.5, 128.1, 128.2, 128.3, 128.6, 129.5,
129.9, 130.5, 130.6, 135.5, 135.7, 144.7, 144.9, 145.9, 159.0, 171.5
(Car). MS (ESI): m/z = 693.2 [M+Na]+, 709.1 [M+K]+.
PAGE purification was performed on 15% or 20% polyacrylamide
gels (acrylamide:bisacrylamide 19:1) in 1ꢁ TBE buffer. Gel dimen-
sions: 0.7 ꢁ 200 ꢁ 320 mm. Samples were loaded with formamide
loading buffer (containing bromophenol blue and xylene cyanol as
dye markers) and run at 35 W. Bands were detected by UV shad-
owing and extracted by the ‘crush and soak’ method using TEN
buffer as previously described.35
4.2.5. 50-O-(40,40-Dimethoxytrityl)-O4-(4-chlorophenyl)-20-O-
methlyuridine 30-(2-cyanoethyl-N,N-
diisoproyplphosphoramidite) (9)
50-O-DMT-O4-(4-chlorophenyl)-20-O-methyluridine
(8)
4.4. In vitro transcription of donor fragments
(93.6 mg, 0.14 mmol, 1.00 equiv) was dried under high vacuum
overnight and dissolved in DCE (2.5 mL). N,N-Dimethylethylamine
(109 mg, 1.40 mmol, 10.0 equiv) and 2-cyanoethyl-(N,N-diisopro-
pylamino) chlorophosphite (57.8 mg, 0.24 mmol, 1.75 equiv) were
added and the mixture stirred for 2 h. The reaction was stopped by
addition of a few drops of methanol, the organic layer was diluted
with DCM and washed with NaHCO3 and water, dried over Na2SO4
and evaporated under reduced pressure. The crude product was
purified by column chromatography on SiO2 with 20–50% ethyl
acetate in hexane (containing 2% NEt3), to yield target compound
9 (80.4 mg, 0.09 mmol, 66%) as white foam. Rf (hexane/ethyl ace-
tate 1:1) = 0.57; 1H NMR (400 MHz, CDCl3): d (ppm) = 0.98 (d,
6 H, J = 6.8 Hz, (CH3)2CHN), 1.11–1.26 (m, 18 H, (CH3)2CHN), 2.39
(t, 2 H, J = 6.0 Hz, CH2CH2CN-Ha), 2.60 (t, 2 H, J = 6.0 Hz,
CH2CH2CN–Hb), 3.42–3.88 (m, 14 H, H-C(20), 2 ꢁ H-C(50),
((CH3)2CH)2N, 2 ꢁ CH2CH2CN), 3.63 (s, 6 H, 20-OCH3), 3.78–3.79
(m, 12 H, OCH3), 4.22–4.26 (m, 2 H, H-C(40)), 4.43 (td, 1 H,
J = 9.0 Hz, J = 4.7 Hz, H-C(30) diastereomer 1), 4.61 (td, 1 H,
J = 9.0 Hz, J = 4.7 Hz, H-C(30) diastereomer 2), 4.48 (d, 1 H,
J = 7.4 Hz, H-C(5), diastereomer 1), 5.54 (d, 1 H, J = 7.4 Hz, H-C(5),
diastereomer 2), 5.96 (s, 1 H, H-C(10), diastereomer 1), 5.99 (s,
1 H, H-C(10), diastereomer 2), 6.80–6.86 (m, 8 H, DMT-Har), 7.05–
7.09 (m, 4 H, pCl-Ph-Har), 7.23–7.43 (m, 22 H, Har), 8.53 (d, 1 H,
J = 7.4 Hz, H-C(6), diastereomer 1), 8.59 (d, 1 H, J = 7.4 Hz, H-C(6),
diastereomer 2); 31P NMR (162 MHz, CDCl3): d (ppm) = 150.1 (s,
diastereomer 1), 150.5 (s, diastereomer 2); MS (ESI): m/z = 893.3
[M+Na]+.
In vitro transcription was performed as in our previous re-
ports35 with T7 RNA polymerase and synthetic DNA templates
(1 lM) using 4 mM of each NTP in 40 mM Tris–HCl, pH 8.0,
30 mM MgCl2, 10 mM DTT, and 2 mM spermidine at 37 °C
for 5 h.
4.5. General procedure for DNA-catalyzed ligation of RNA
fragments
Spin-labeled or unmodified acceptor RNA, in vitro transcribed
donor RNA, and 9DB1⁄ DNA were combined in equimolar ratio
and annealed in 25 mM HEPES, pH 7.5, 15 mM NaCl, 0.1 mM EDTA
by heating to 95 °C for 2 min and slow cooling to room tempera-
ture. The reaction buffer and divalent metal ions were added from
5ꢁ concentrated stock solutions. The final concentrations were
50 mM HEPES, pH 7.5, 150 mM NaCl, 2 mM KCl, 20 mM MnCl2 or
50 mM CHES, pH 9.0, 150 mM NaCl, 2 mM KCl, 40 mM MgCl2. Liga-
tion reactions were performed on a scale of 150 pmol up to 3 nmol
per substrate at a final concentration of each strand at 15 lM. The
ligation mixture was incubated at 37 °C for 5 h (or at 25 °C for 12 h
for the shortest substrates). EDTA was added (50 mM) and the RNA
and DNA strands were precipitated by addition of cold ethanol
(3 vol). Reactions were analyzed, and products were isolated by an-
ion exchange HPLC or PAGE.
4.6. Ligation of RNA by T4 DNA ligase
4.3. Synthesis of spin-labeled RNA
Equimolar amounts of acceptor RNA, 50-phosphorylated donor
RNA and DNA splint were annealed by heating to 95 °C for 2 min
and slow cooling to ambient temperature (final RNA concentration
RNA solid-phase syntheses were performed on a Pharmacia
Gene Assembler Plus, using 0.7 lmol polystyrene custom primer
support from GE Healthcare. 20-O-TOM-protected ribonucleoside
phosphoramidites were used for all unmodified RNA nucleotide
positions (100 mM in CH3CN). Convertible nucleoside phospho-
ramidites were dried over-night under high vacuum and then dis-
solved in dry CH3CN. The solutions were dried over molecular
sieves for several hours before installation on the gene assembler.
S-Benzylthiotetrazole (BTT) (250 mM in CH3CN) was used as acti-
vator; coupling time was 4 min. Detritylation, capping and oxida-
tion was performed under standard conditions.
was 15
tion 40 mM Tris–HCl, pH 7.8, 10 mM MgCl2, 10 mM DTT, 0.5 mM
ATP) and T4 DNA ligase (final concentration 0.5 U/ L) were added,
lM of each strand). T4 DNA ligase buffer (final concentra-
l
and the reaction mixture was incubated at 37 °C for 6–12 h and
analyzed by anion exchange HPLC or PAGE.
4.7. Kinetic assays of DNA-catalyzed ligation
Solid support containing the fully protected RNA (ca. 0.3
was incubated with 4-amino-TEMPO or 3-amino-proxyl (100
2 M in methanol) for 24–48 h at 42–55 °C. For all reactions with
3-amino-proxyl and for all RNAs longer than 15 nt, MeNH2 in
H2O (500 lL, 8 M) or NH4OH (500 lL, 25% aq) was added and incu-
bated in a tightly closed vial for 2–5 h (37 °C for MeNH2, 55 °C for
NH4OH). The solid support was removed by filtration, and the sol-
l
mol)
L,
Assays of 9DB1⁄-catalyzed ligation kinetics were performed un-
der single-turnover conditions with 50-32P-labeled acceptor RNA as
described.35 Time points were taken up to 5 h, and samples were
analyzed by PAGE. Gels were dried, exposed to a Phosphor storage
screen, and imaged with a PhosphorImager. The ligation yield was
determined by volume integration and the data were fit to the
equation yield = Y⁄(1 – eꢂkt), where k = kobs and Y = final yield.
l

6180
L. Büttner et al. / Bioorg. Med. Chem. 21 (2013) 6171–6180
8. Sicoli, G.; Wachowius, F.; Bennati, M.; Höbartner, C. Angew. Chem., Int. Ed. 2010,
4.8. Thermal melting experiments
RNA duplexes were prepared at 1–25
49, 6443.
9. Höbartner, C.; Sicoli, G.; Wachowius, F.; Gophane, D. B.; Sigurdsson, S. T. J. Org.
Chem. 2012, 77, 7749.
l
M in 10 mM potassium
10. Roth, A.; Breaker, R. R. Annu. Rev. Biochem. 2009, 78, 305.
11. Serganov, A.; Nudler, E. Cell 2013, 152, 17.
phosphate buffer, pH 7.0, 150 mM NaCl. Melting curves were mon-
itored in quartz cuvettes of 1 cm or 0.1 cm path length at 250, 260,
270, and 280 nm on a Cary 100 UV spectrophotometer (Varian Inc.)
equipped with a multiple cell holder and a Peltier temperature-
controller. Two full heating and cooling cycles (4 ramps) were col-
lected with a heating/cooling rate of 0.7 °C/min. Hyperchromicity
at 250 nm was calculated, and the Tm was determined at the max-
imum of the first derivative of the hyperchromicity curve. Thermo-
dynamic parameters were obtained by analysis of ln(cT) versus
1/Tm, with cT calculated from A260 at Tm.
12. Haller, A.; Souliere, M. F.; Micura, R. Acc. Chem. Res. 2011, 44, 1339.
13. Frilander, M. J.; Turunen, J. J. In Handbook of RNA Biochemistry; Hartmann, R. K.,
Bindereif, A., Schön, A., Westhof, E., Eds.; Wiley-VCH, 2008; p 36.
14. Haller, A.; Altman, R. B.; Souliere, M. F.; Blanchard, S. C.; Micura, R. Proc. Natl.
Acad. Sci. U.S.A. 2013, 110, 4188.
15. Haller, A.; Rieder, U.; Aigner, M.; Blanchard, S. C.; Micura, R. Nat. Chem. Biol.
2011, 7, 393.
16. Rieder, R.; Höbartner, C.; Micura, R. Methods Mol. Biol. 2009, 540, 15.
17. Lang, K.; Micura, R. Nat. Protoc. 2008, 3, 1457.
18. Brenner, M. D.; Scanlan, M. S.; Nahas, M. K.; Ha, T.; Silverman, S. K. Biochemistry
2010, 49, 1596.
19. Moore, M. J.; Sharp, P. A. Science 1992, 256, 992.
20. Silverman, S. K.; Cech, T. R. Biochemistry 1999, 38, 14224.
21. Höbartner, C.; Silverman, S. K. Angew. Chem., Int. Ed. 2005, 44, 7305.
22. Dönmez, G.; Hartmuth, K.; Kastner, B.; Will, C. L.; Lührmann, R. Mol. Cell 2007,
25, 399.
23. Getz, E. B.; Xiao, M.; Chakrabarty, T.; Cooke, R.; Selvin, P. R. Anal. Biochem. 1999,
273, 73.
24. Hyodo, F.; Matsumoto, K.; Matsumoto, A.; Mitchell, J. B.; Krishna, M. C. Cancer
Res. 2006, 66, 9921.
25. Paletta, J. T.; Pink, M.; Foley, B.; Rajca, S.; Rajca, A. Org. Lett. 2012, 14, 5322.
26. Marx, L.; Chiarelli, R.; Guiberteau, T.; Rassat, A. J. Chem. Soc., Perkin Trans. 1
2000, 1181.
27. Krstic, I.; Hänsel, R.; Romainczyk, O.; Engels, J. W.; Dötsch, V.; Prisner, T. F.
Angew. Chem., Int. Ed. 2011, 50, 5070.
4.9. EPR experiments
CW-EPR spectra were acquired on a Bruker EMX spectrometer
at 22 °C, using 100 kHz modulation frequency, 0.8 G modulation
amplitude, and 2.0 mW microwave power. Measurements were ta-
ken in quartz capillaries containing 40 lM or 200 lM spin-labeled
RNA in 2.5 mM potassium phosphate buffer, pH 7.0, 30 mM NaCl.
The duplex samples contained 1.5 equiv of the unlabeled comple-
mentary strand. Double strands were annealed by heating to
95 °C for 2 min and slow cooling to room temperature before
transfer to the capillary.
28. Azarkh, M.; Singh, V.; Okle, O.; Seemann, I. T.; Dietrich, D. R.; Hartig, J. S.;
Drescher, M. Nat. Protoc. 2013, 8, 131.
29. Tkach, I.; Sicoli, G.; Höbartner, C.; Bennati, M. J. Magn. Reson. 2011, 209, 341.
30. Tkach, I.; Pornsuwan, S.; Höbartner, C.; Wachowius, F.; Sigurdsson, S. T.;
Baranova, T. Y.; Diederichsen, U.; Sicoli, G.; Bennati, M. Phys. Chem. Chem. Phys.
2013, 15, 3433.
31. Egli, M.; Pallan, P. S. Annu. Rev. Biophys. Biomol. Struct. 2007, 36, 281.
32. Schlosser, K.; Li, Y. Chem. Biol. 2009, 16, 311.
33. Höbartner, C.; Pradeepkumar, P. I. In New Strategies in Chemical Synthesis and
Catalysis; Pignataro, B., Ed.; Wiley-VCH, 2012; p 129.
34. Purtha, W. E.; Coppins, R. L.; Smalley, M. K.; Silverman, S. K. J. Am. Chem. Soc.
2005, 127, 13124.
35. Wachowius, F.; Javadi-Zarnaghi, F.; Höbartner, C. Angew. Chem., Int. Ed. 2010,
49, 8504.
Acknowledgments
We gratefully acknowledge financial support by the DFG (IRTG
1422 and GSC 226/1) and the Max Planck Society. A.O. thanks the
Beilstein foundation for a research scholarship. We thank Professor
Dr. Marina Bennati, MPIbpc for access to the Bruker EMX spec-
trometer, and Drs. Soraya Pornsuwan and Igor Tkach for assistance
with EPR data acquisition. Uwe Plessmann (Facility for Bioanalyti-
cal Mass spectrometry) and Jürgen Bienert (Facility for Synthetic
Chemistry) are acknowledged for recording MS and NMR spectra.
36. Pitsch, S.; Weiss, P. A.; Jenny, L.; Stutz, A.; Wu, X. Helv. Chim. Acta 2001, 84,
3773.
37. McGee, D. P.; Vaughn-Settle, A.; Vargeese, C.; Zhai, Y. J. Org. Chem. 1996, 61,
781.
Supplementary data
38. Höbartner, C.; Micura, R. J. Am. Chem. Soc. 2004, 126, 1141.
39. McGee, D. P. C.; Zhai, Y. S. Nucleosides Nucleotides 1996, 15, 1797.
40. Kloiber, K.; Spitzer, R.; Tollinger, M.; Konrat, R.; Kreutz, C. Nucleic Acids Res.
2011, 39, 4340.
41. Lu, C.; Smith, A. M.; Fuchs, R. T.; Ding, F.; Rajashankar, K.; Henkin, T. M.; Ke, A.
Nat. Struct. Mol. Biol. 2008, 15, 1076.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2013.04.007.
42. Heppell, B.; Lafontaine, D. A. Biochemistry 2008, 47, 1490.
43. Heppell, B.; Blouin, S.; Dussault, A. M.; Mulhbacher, J.; Ennifar, E.; Penedo, J. C.;
Lafontaine, D. A. Nat. Chem. Biol. 2011, 7, 384.
44. Wunnicke, D.; Strohbach, D.; Weigand, J. E.; Appel, B.; Feresin, E.; Suess, B.;
Müller, S.; Steinhoff, H.-J. RNA 2011, 17, 182.
References and notes
1. Gesteland, R. F.; Cech, T. R.; Atkins, J. F. The RNA World, 3rd ed.; CSHL Press,
2006.
2. Wachowius, F.; Höbartner, C. ChemBioChem 2010, 11, 469.
3. Solomatin, S.; Herschlag, D. Methods Enzymol. 2009, 469, 47.
4. Shelke, S. A.; Sigurdsson, S. T. Eur. J. Org. Chem. 2012, 2291.
5. Piton, N.; Mu, Y.; Stock, G.; Prisner, T. F.; Schiemann, O.; Engels, J. W. Nucleic
Acids Res. 2007, 35, 3128.
6. Qin, P. Z.; Haworth, I. S.; Cai, Q.; Kusnetzow, A. K.; Grant, G. P. G.; Price, E. A.;
Sowa, G. Z.; Popova, A.; Herreros, B.; He, H. Nat. Protoc. 2007, 2, 2354.
7. Allerson, C. R.; Chen, S. L.; Verdine, G. L. J. Am. Chem. Soc. 1997, 119, 7423.
45. Krstic, I.; Frolow, O.; Sezer, D.; Endeward, B.; Weigand, J. E.; Suess, B.; Engels, J.
W.; Prisner, T. F. J. Am. Chem. Soc. 2010, 132, 1454.
46. Zlatev, I.; Lackey, J. G.; Zhang, L.; Dell, A.; McRae, K.; Shaikh, S.; Duncan, R. G.;
Rajeev, K. G.; Manoharan, M. Bioorg. Med. Chem. 2013, 21, 722.
47. Zlatev, I.; Lavergne, T.; Debart, F.; Vasseur, J. J.; Manoharan, M.; Morvan, F. Org.
Lett. 2010, 12, 2190.
48. Nguyen, P.; Qin, P. Z. Wiley Interdiscip. Rev. RNA 2012, 3, 62.