A chemo-enzymatic approach to specifically click-modified RNA

Article abstract of DOI:10.1039/c3cc40594j

The growing interest in single-molecule analysis of RNA calls for programmable enzymatic labeling strategies beyond the horizon of solid-phase synthesized RNAs. Herein we describe an easy and versatile chemo-enzymatic approach to label RNA at its termini or defined internal positions via click-chemistry.

Full text of DOI:10.1039/c3cc40594j

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
A chemo-enzymatic approach to specifically
click-modified RNA†
Cite this: Chem. Commun., 2013,
49, 3128
Claudine M. Dojahn, Marlen Hesse and Christoph Arenz*
Received 23rd January 2013,
Accepted 13th February 2013
DOI: 10.1039/c3cc40594j
www.rsc.org/chemcomm
The growing interest in single-molecule analysis of RNA calls for in length.3 Paredes and Das picked up and transferred this idea to
programmable enzymatic labeling strategies beyond the horizon in vitro-transcribed RNA oligonucleotides that were successfully
of solid-phase synthesized RNAs. Herein we describe an easy and equipped with azido groups at the 50-terminus by adding an
versatile chemo-enzymatic approach to label RNA at its termini or appropriate azido-transcription starter. Moreover they addressed
defined internal positions via click-chemistry.
the 30-terminus enzymatically with a 30-azido-20,30-dideoxytriphos-
phate and a poly(A) polymerase. Afterwards they conjugated those
After sequencing the human genome, the analysis of RNA-related oligonucleotides with synthetic RNAs modified with commercially
processes like gene regulation by micro RNAs (miRNAs) as well as available alkynylated reagents.6
their utilization in therapeutic strategies has attracted increasing
For internal labelling nucleoside triphosphates (NTPs) equipped
attention.1 For more detailed investigations of their mode of action with alkynes are available but are randomly incorporated by RNA
and to overcome the limitations of effective delivery to the targeted polymerases, while the labelling efficiency depends on the chemical
cells, powerful/suitable/new RNA-labeling strategies are required.2
nature and the bulkiness of the probe to be incorporated.4,10 An
Hitherto, most of the modifications in RNA are introduced by example for a site-specific internal labeling approach was reported
solid-phase synthesis which is struggling with lower coupling by Winz et al., who installed 20-azido-nucleoside triphosphates at the
efficiencies and more complex purification protocols in comparison 30-terminus of RNA by using a poly(A) polymerase.7 After selection of
to DNA. Therefore the synthesis of RNA oligonucleotides is more the appropriate enzyme and under optimised conditions, only one
expensive, time-consuming, laborious and usually reaches its limits nucleotide was attached. The design of this terminal nucleotide
at sequence lengths longer than 50 nucleotides.3 Enzymatic enabled a strand prolongation via a splinted ligation, shifting the
methods, e.g. in vitro-transcriptions or ligations of short RNA former terminal nucleotide to an internal position. Herein we
strands, provide an attractive solution to these problems.4
present a powerful method for chemo-enzymatic synthesis of term-
Both, solid-phase and enzymatic RNA synthesis allow incorpora- inally and internally labelled RNA, without the need for tedious
tion of modified (ribo)nucleotides. In cases of incompatibilities due optimisation and splint ligation.
to reactive reagents or a low substrate tolerance, a small and
Our intention to detect Dicer-mediated miRNA maturation
functionalisable group is used instead and is attached with encouraged us to label its hairpin-shaped precursors (pre-miRNA)
the desired modification post-synthetically. Referring to this, the at both ends with a fluorophore and a quencher, respectively, to
1,3-diploar Cu-catalyzed azide–alkyne cycloaddition (CuAAC) has quantify its Dicer-mediated cleavage by measuring the increasing
proven to be a powerful method. However this is primarily the case fluorescence over time, which results from the absence of
for DNA oligonucleotides because RNA is more sensitive to copper- quenching.11,12 Due to the known limitations of solid-phase
induced strand-cleavage.5 Although this effect can be repressed by synthesis and in order to establish a method that allows a
the addition of copper-stabilizing ligands only a few reports have modular conjugation of different labels, we envisioned a tech-
described modification of RNA by means of CuAAC.3,6–9
nique based on CuAAC in combination with enzymatic methods.
To introduce an alkyne group at the 50-end of RNA we developed
Using synthetic RNA strands bearing alkyne and azido groups
which are suitable for cross-strand and intrastrand CuAAC-ligations, a short and efficient route to novel O-propargyl-guanosine-50-mono-
El-Sagheer and Brown have assembled DNA–RNA hybrids B100 nt phosphate. In vitro transcriptions in the presence of this starter
provide a 50-terminal alkyne modified RNA molecule that is ready
for click-reactions (Fig. 1 and 2).13,14 A tetrazol-mediated one-pot-
phosphoramidite-coupling of TMS-protected propargyl alcohol and
the protected 50-hydroxylguanosine followed by removal of the
protecting groups afforded the desired compound 9 (GMPPrg) as a
¨
Humboldt Universitat zu Berlin, Institute for Chemistry, Brook-Taylor-Str. 2,
12489 Berlin, Germany. E-mail: christoph.arenz@chemie.hu-berlin.de;
Fax: + 49 302093-6947
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3cc40594j
c
3128 Chem. Commun., 2013, 49, 3128--3130
This journal is The Royal Society of Chemistry 2013

View Article Online
Communication
ChemComm
Fig. 1 (a) Ac2O, 1,4-dioxan; (b) (NH4)2Ce(NO3)6, I2, MeCN, 80 1C; (c) NEt3, CuI, PdCl2(PPh3)2, DMF; d1 TBAF, MeCN, d2 EtOH, NaOH; (e) tBu2Si(OTf)2, TBDMS-Cl, imidazole;
(f) HFꢀpyridine, THF; (g) P(OCE)2(NiPr2), tetrazole, CH2Cl2, tBuOOH; h1 BSA, DBU, pyridine, h2 NEt3ꢀ3HF; (i) acetone, HClO4; (j) N,N-di-methylformamide dimethylacetal,
DMF; k1 P(OCE)(NiPr2)2, tetrazole, CH2Cl2 k2 tetrazole, 3-trimethylsilyl propargyl alcohol k3 tBuOOH; (l) 7 N NH3 in MeOH; (m) TFA–H2O 1 : 1, TEAB buffer.
For labelling the 30-terminus we envisioned the use of
commercially available T4 RNA ligase 1 (T4) which ligates RNA
strands or attaches singular 30,50-bisphosphates to the 30-end.18,19
Blocking the 30-hydroxyl group with a phosphate prevents multiple
incorporation, however after completion of the ligation, the terminal
phosphate can be easily removed by Calf Intestinal Phosphatase
(CIP). To address the 30-terminus we established a synthesis route to
alkynylated 30,5-bisphosphate uridine 5 (pUpAlk), which has not been
described so far. The synthesis of this building block started from
uridine 1 which was modified with an ethinyl group at the
C5-position by known procedures (total yield 51%).20 By using a
silyl-based protecting group strategy, the key compound 2 was
Fig. 2 Illustration of the 50-, 30-end labelling methodology.
triethylamine salt in 17% yield over all steps.15 In the following regioselectively phosphorylated.21 Removal of the cyanoethyl groups
in vitro-transcription with T7 polymerase (T7), the building block 9 failed first when using methanolic ammonia solution but finally
competed with unmodified guanosine triphosphate (GTP) for the succeeded when switching to bis(trimethylsilyl)acetamide and DBU
incorporation at the initiation site of the growing pre-miRNA- in pyridine.22 After silyl deprotection, the azido labelled FAM was
transcript. To favour the implementation of 9 without inducing conjugated by means of CuAAC (ESI†) and ligated to the
transcription abort we integrated the so-called fed-batch-approach DABpre-miRNAs with T4 in line with the manufacturer’s protocol.
into the manufacturer’s protocol.16 Accordingly we started the The surplus of pUpFAM was separated with centrifugal filters and a
in vitro-transcription with a 9-fold excess of GMPPrg over GTP but final HPLC purification showed no leftovers of non-ligated
added further equivalents of GTP every 15 min until the starting DABpre-miRNAs, suggesting a complete conversion of the substrate.
concentration of (A/U/C)TP was reached. After digestion of the DNA Adding the pre-miRNA processing enzyme Dicer to each of those
template and PCI-extraction the mixture of GTP and GMPPrg-primed constructs results in an increase in fluorescence (lem 520 nm),
RNA-transcripts was precipitated with EtOH and used without any confirming that these labelled pre-miRNAs are tolerated as substrates
further purification steps in a subsequent CuAAC. This reaction was (Fig. S4, ESI†). Moreover, all probes were analysed using denatured
carried out in PUS-buffer with DABCYL-azide 11, copper sulfate, PAGE. As expected all bands showed fluorescent activity in the
sodium ascorbate and THPTA as ligand.17 After 2 h, HPLC purifica- absence of SYBRs gold and prove the formation of the full-length
tion enabled us to easily separate the DABCYL-conjugated RNA from product. Molecular masses were confirmed by UHPLC-UV/Vis-MS
the GTP-primed transcript. No residues of RNA starting with the (see ESI†). The strength of this chemo-enzymatic approach lies in its
initiator GMPPrg were detected, indicating that the CuAAC was highly effectiveness and easy handling as we used standard protocols and
efficient (ESI†). Without any further optimizations for the transcrip- commercially available enzymes. Moreover the chosen T4-ligation
tion in terms of salt concentrations, temperature, enzyme mutants method does not need to be optimized in regard to multiple single-
etc. 11 was linked to the 50-end of four different DABCYL-labelled nucleotide-attachments and does not require a template.
pre-miRNAs (pre-let7, pre-miRNA-21, 122, 142,) with overall labeling-
efficiencies between 21 and 38%.
The fluorophore FAM is not the best choice when it comes to
experiments in the cell because its maximum of emission lies
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 3128--3130 3129

View Article Online
ChemComm
Communication
labelled with FAM the elongated pre-miRNA-122 showed no label-
ling. This is possibly due to steric hindrance of the alkyne position
caused by duplex formation in the pre-miRNA-122 whereas the
alkyne group in pre-let7 is easy accessible. A similar observation
7,28
¨
was made by Jaschke et al. (RNA) and Carell et al. (DNA).
In summary we described for the first time the synthesis of
alkynylated 30,50-bisphosphate uridine 5 and O-propargyl-guanosin-
50-monophosphate 9. These building blocks enabled us to establish
a chemoenzymatic approach to label RNA at its termini with
organic molecules like FAM and Dabcyl bearing an azido group.
Additionally, the scope of this methodology was extended to label
RNA site specifically at internal positions. In general, this method
has proven to be very effective and easy to apply what has been
exemplified by the synthesis of several pre-miRNAs probes up to
79 nt in length. The chemoenzymatic approach suggests that much
bigger RNA molecules are also feasible.
Fig. 3 Internal labelling of extended pre-miRNAs by consecutive ligation–
dephosphorylation: left: extended pre-let7, right: extended pre-miR-122A.
in the range of the cellular auto-fluorescence and it is known to be
sensitive to photobleaching.23,24 To tackle this problem we chose
Alexa Fluors 633 as an alternative, however when we started this
project there was no azido functionalized Alexa Fluors 633
purchasable. For this reason, we modified the established strategy
to label the biologically highly relevant pre-miR-122. Instead of
ligating a single nucleotide to the 30-end we decided to build up
pre-miRNA-122 from two oligonucleotide fragments (Fig. S5a,
ESI†). The DABCYL-tagged 50-end was synthesized according to
the aforementioned protocol. The remaining Alexa-RNA strand
from the 30-terminus was obtained from a commercial source and
synthesized by solid-phase synthesis. A T4-ligation of both
strands, again without any template or splint, led to the full-
length product in almost quantitative yield (Fig. S5c, ESI†). The
HPLC purified probe was incubated with Dicer and this led to a
5-fold enhancement of fluorescence within 4 hours whereas the
treatment with denatured Dicer showed almost no increase.
While several methods for terminal labelling of RNA exist,
site-specific internal labelling of RNA is a special challenge.7,25–27
Our work on the Alexa-labelled pre-miRNA inspired us to extend
the scope of our ligation strategy based on the novel building
block pUpAlk. The 30-phosphate of the UAlk-elongated 30end can
be removed with CIP thereby creating a new ligation site which
could be elongated with a further RNA strand. This concept
Notes and references
1 C. Arenz, Angew. Chem., Int. Ed., 2006, 45, 5048–5050.
2 K. Phelps, A. Morris and P. A. Beal, ACS Chem. Biol., 2011, 7,
100–109.
3 H. El-Sagheer and T. Brown, Proc. Natl. Acad. Sci. U. S. A., 2010, 107,
15329–15334.
4 Y. Jao and A. Salic, Proc. Natl. Acad. Sci. U. S. A., 2008, 105,
15779–15784.
5 H. El-Sagheer and T. Brown, Chem. Soc. Rev., 2010, 39, 1388–1405.
6 E. Paredes and S. R. Das, ChemBioChem, 2011, 12, 125–131.
¨
7 M.-L. Winz, A. Samanta, D. Benzinger and A. Jaschke, Nucleic Acids
Res., 2012, 40, e78.
8 T. Yamada, C. G. Peng, S. Matsuda, H. Addepalli, K. N. Jayaprakash,
M. R. Alam, K. Mills, M. A. Maier, K. Charisse, M. Sekine,
M. Manoharan and K. G. Rajeev, J. Org. Chem., 2011, 76, 1198–1211.
9 K. Fauster, M. Hartl, T. Santner, M. Aigner, C. Kreutz, K. Bister,
E. Ennifar and R. Micura, ACS Chem. Biol., 2012, 7, 581–589.
10 M. Grammel, H. Hang and N. K. Conrad, ChemBioChem, 2012, 13,
1112–1115.
11 P. Davies and C. Arenz, Angew. Chem., Int. Ed., 2006, 45, 5550–5552.
12 P. Davies and C. Arenz, Bioorg. Med. Chem., 2008, 16, 49–55.
13 J. F. Milligan, D. R. Groebe, G. W. Witherell and O. C. Uhlenbeck,
Nucleic Acids Res., 1987, 15, 8783–8798.
14 J. F. Milligan and O. C. Uhlenbeck, Methods Enzymol., 1989, 180,
51–62.
15 L. Zhang, L. Sun, Z. Cui, R. L. Gottlieb and B. Zhang, Bioconjugate
Chem., 2001, 12, 939–948.
16 J. A. Kern and R. H. Davis, Biotechnol. Prog., 1999, 15, 174–184.
´
¨
17 R. Fiammengo, K. Musılek and A. Jaschke, J. Am. Chem. Soc., 2005,
127, 9271–9276.
would provide a smart alternative to the few methods known so 18 J. R. Barrio, M. d. C. G. Barrio, N. J. Leonard, T. E. England and
O. C. Uhlenbeck, Biochemistry, 1978, 17, 2077–2081.
19 T. E. England and O. C. Uhlenbeck, Nature, 1978, 275, 560–561.
20 S. M. Park, Y. Shen and B. H. Kim, Org. Biomol. Chem., 2007, 5,
far for the site-specific internal labelling of RNAs which require
laborious optimizations or template modifications.7,25–27
To verify this idea we ligated pUpAlk to pre-let7. Instead of
HPLC purification we used centrifugal filters to remove any
excess of pUpAlk. Analytical UHPLC-UV/Vis-MS analysis of the
crude product verified a complete conversion of the unlabelled
pre-let7 illustrating the effectiveness of this reaction. In the next
step phosphate groups were removed and the deblocked
30-hydroxyl function used for a T4-ligation of a second RNA
strand. Finally we carried out a CuAAC together with FAM-N3.
The same procedure was applied to a shortened pre-miRNA-122.
Adding the second RNA strand to the UAlk-prolonged RNA
provides the natural pre-miRNA-122 enlarged by one internal
alkynylated uridine (Fig. 3). A PAGE-analysis is shown in Fig. S6,
ESI.† In contrast to the elongated pre-let7 which was successfully
610–612.
21 V. Serebryany and L. Beigelman, Tetrahedron Lett., 2002, 43,
1983–1985.
22 M. Sekine, H. Tsuruoka, S. Iimura and T. Wada, Nat. Prod. Lett.,
1994, 5, 41–46.
¨
23 R. Sjoback, J. Nygren and M. Kubista, Spectrochim. Acta, Part A, 1995,
51, L7–L21.
24 L. Song, E. J. Hennink, I. T. Young and H. J. Tanke, Biophys. J., 1995,
68, 2588–2600.
25 Y. Motorin, J. Burhenne, R. Teimer, K. Koynov, S. Willnow,
E. Weinhold and M. Helm, Nucleic Acids Res., 2010.
ˇ
´
26 M. Tomkuviene, B. Clouet-d’Orval, I. Cerniauskas, E. Weinhold and
ˇ
S. Klimasauskas, Nucleic Acids Res., 2012, 40, 6765–6773.
27 T. Ishizuka, M. Kimoto, A. Sato and I. Hirao, Chem. Commun., 2012,
48, 10835–10837.
28 T. Wirges, P. M. E. Gramlich, K. Gutsmiedl, J. Gierlich, G. A. Burley
and T. Carell, QSAR Comb. Sci., 2007, 26, 1159–1164.
c
3130 Chem. Commun., 2013, 49, 3128--3130
This journal is The Royal Society of Chemistry 2013