Identification of efficient and sequence specific bimolecular artificial ribonucleases by a combinatorial approach

Article abstract of DOI:10.1039/b712532a

Chemically modified nucleotide monomers were incorporated into adjacent terminal positions of two separate oligonucleotides complementary to an RNA target; all possible combinations of the catalytic units were tested, resulting in an artificial nuclease t

Full text of DOI:10.1039/b712532a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Identification of efficient and sequence specific bimolecular artificial
ribonucleases by a combinatorial approach{{
Marcin Kalek,§a Peter Benediktson,a Birte Vesterb and Jesper Wengel*a
Received (in Cambridge, UK) 3rd September 2007, Accepted 26th November 2007
First published as an Advance Article on the web 12th December 2007
DOI: 10.1039/b712532a
Chemically modified nucleotide monomers were incorporated
into adjacent terminal positions of two separate oligonucleo-
tides complementary to an RNA target; all possible combina-
tions of the catalytic units were tested, resulting in an artificial
nuclease that showed high activity and catalytic turnover.
There is a growing interest in artificial systems capable of sequence
specific RNA cleavage as an alternative to RNAi, RNaseH,
ribozyme and DNAzyme approaches for antisense therapy, and as
general tools for RNA manipulation.1–5 Artificial agents for DNA
cleavage have also been developed.6 In this paper, we present a
novel combinatorial method that allows efficient identification and
optimization of artificial ribonucleases, herein defined as oligonu-
Fig. 1 Designs of artificial ribonucleases.
cleotides being conjugated with non-nucleotide moieties facilitating
RNA cleavage.
probably because a productive balance between nuclease flexibility,
RNA target affinity and product release was reached.7
The majority of sequence specific artificial ribonucleases
discovered so far include an oligonucleotide unit to ensure target
Rational design of artificial nucleases is inherently difficult and
recognition, and an attached catalytic group able to mediate RNA
we therefore believe that a combinatorial approach would be the
phosphodiester backbone hydrolysis. Most attention has been
ultimate solution in the search for efficient systems. Inspired by the
devoted to metal complexes as catalytic groups which have also
above mentioned work we adopted the separate dinuclear artificial
proven most efficient due to the fact that phosphodiester
nuclease setup as a base of our screening strategy. The approach
hydrolysis can be assisted by metal ions in several distinct catalytic
relies on systematic testing of different constitutions of catalytic
pathways.2,5 Although this versatility makes metal-dependent
monomers incorporated into adjacent terminal positions of the
mononuclear artificial nucleases (Fig. 1 A) relatively efficient, its
two individual oligonucleotide strands (DNA1 and DNA2; Fig. 2)
full potential is likely to be exploited only with dinuclear systems
(Fig. 1 B–D). The cooperative action of two catalytic species has
accordingly been reported to lead to a remarkable increase in
phosphodiester cleavage activity.7
In the standard design of a dinuclear artificial nuclease (Fig. 1 B)
two catalytic units are incorporated into a single oligonucleotide.8
Recently, it was shown that introduction of a certain degree of
flexibility into such system may result in increased cleavage
efficiency7 as demonstrated in an initial artificial nuclease design
involving two oligonucleotides, each possessing a Cu(II) complex
(Fig. 1 C). Upon interconnection of these two oligonucleotides by
a flexible linker (Fig. 1 D) RNA cleavage capability increased,
aThe Nucleic Acid Center, Department of Physics and Chemistry,
University of Southern Denmark, DK-5230, Odense M, Denmark.
E-mail: jwe@ifk.sdu.dk; Fax: +45 6550 4385; Tel: +45 6550 2510
bThe Nucleic Acid Center, Department of Biochemistry and Molecular
Biology, University of Southern Denmark, DK-5230, Odense M,
Denmark
{ Electronic supplementary information (ESI) available: Synthesis proce-
dures; experimental protocols for oligonucleotide synthesis and purifica-
tion; details of RNA cleavage reactions. See DOI: 10.1039/b712532a
{ The HTML version of this article has been enhanced with colour
Fig. 2 Sequences of the bimolecular artificial nucleases and the RNA
images.
target. The arrows indicate cleavage sites. Also shown are structures of the
monomers used in this study.
§ Current affiliation: Department of Organic Chemistry, Arrhenius
Laboratory, Stockholm University.
762 | Chem. Commun., 2008, 762–764
This journal is ß The Royal Society of Chemistry 2008

View Article Online
by systematically combining these in all possible pairs. Such
methodology is much more convenient than the preparation and
screening of a standard dinuclear artificial nuclease library
(Fig. 1 B), since it requires the synthesis of fewer oligonucleotides
(2n vs. n2, possessing n different RNA cleaving units).
For this study four novel conjugates of nucleosides with metal
complexing monomers were synthesized and incorporated into
oligonucleotides DNA1 and DNA2 (Fig. 2).{ Three of these
conjugates contain terpyridine, a metal chelator widely used in
artificial nucleases,1,2 whereas the fourth contains a N,N-bis(2-
pyridylmethyl)glycyl moiety.5 The monomers differ in sugar
puckering as the furanose ring in 29-amino-DNA monomers (as
in X, Y and P) is known to adopt an S-type conformation9 and in
29-amino-LNA monomers (as in Z) an N-type conformation.10,11
DNA1 and DNA2 contain, in addition to the metal complexing
monomers, two LNA monomers to ensure good RNA substrate
affinity.11 It should be noted that we have earlier demonstrated
efficient RNA targeting for oligonucleotides composed of DNA,
LNA and functionalized 29-amino-DNA monomers.12
Fig. 4 Dependence of DNA1-Y : DNA2-Y activity on Cu2+ amount
(cDNA1-Y = cDNA2-Y = cRNA = 1 mM; 37 uC). During thermal denaturation
experiments only a single transition was observed due to similar Tm values
of the DNA1-Y : RNA and DNA2-Y : RNA duplexes.
after the cleavage of non-labeled RNA was completed. The MS
spectra (Fig. 5) revealed, in addition to DNA1-Y and DNA2-Y,
the presence of an RNA 59-fragment (9-mer) bearing cyclic
phosphate at its 39 end, together with a 39-fragment (8-mer
possessing a free 59-OH group) as the major species. These
products directly indicate that RNA scission proceeds by a
hydrolytic cleavage mechanism.2 It means that the Cu2+ complex
(or the two complexes) acts as a Lewis acid activating 29-OH, the
phosphate moiety, the leaving group and/or a water molecule.5
The higher cleavage efficiency in the presence of only one Cu2+
equivalent indicates that a free terpyridine moiety is able to assist
during the reaction, for instance by acting as a hydrogen bond
acceptor that aids in forming the optimal catalytic center. In this
context it should be noted that we do not anticipate the saturated
Cu(terpy)2 complex to be catalytically active, and if formed, that it
will be in an equilibrium with the ‘open’ form.
The cleavage efficiency of all sixteen oligonucleotide combina-
tions (4 6 4) was assessed by incubating with complementary
32P-labelled 17-mer RNA target. The RNA cleavage reactions
were analyzed by gel electrophoresis (Fig. 3) and quantified by
phosphor imager scanning. The screening was carried out under
single turnover conditions with a DNA1 : DNA2 : RNA ratio of
4 : 4 : 1 in the presence of excess of Cu2+. The tandem YY
combination appeared to be the most efficient cleaving over 80%
of the target RNA. Scission predominantly occurred between the
nucleotides positioned opposite to the catalytic monomers, and to
less extent in the adjacent positions. DNA1-Y and DNA2-Y when
tested individually under similar conditions cleaved only 23% and
16% of the target, respectively, which proves a synergistic action of
the two catalytic units in the experiment involving both DNA1-Y
and DNA2-Y.{
Next, we investigated cleavage by the most effective combina-
tion with respect to the amount of Cu2+. Surprisingly, maximum
activity was observed in the presence of approximately one
equivalent of Cu2+ (Fig. 4). Adding more Cu2+ decreased the level
of cleavage and a plateau was reached when two or more Cu2+
equivalents were used. As virtually identical thermal stabilities were
measured with one and five Cu2+ equivalents added,{ the observed
difference in cleavage efficiency is not related to target binding
affinity. Instead the results suggest that two distinct mechanisms
are operating in the presence of one and two Cu2+ ions,
respectively, the former cleaving more efficiently than the latter.
To obtain deeper insight into the character of phosphodiester
bond scission, MS analysis of the reaction mixture was carried out
Finally, we evaluated cleavage over time for the optimized
artificial nuclease system at different ratios to target RNA (Fig. 6).
Under single turnover conditions (4 : 4 : 1 and 2 : 2 : 1), the target
Fig. 5 MALDI mass spectra of the cleavage reaction of unlabelled RNA
with DNA1-Y : DNA2-Y, in presence of Cu2+ (cDNA1-Y = cDNA2-Y
=
Fig. 3 Denaturating gel electrophoresis of the screening cleavage
reactions with different combinations of monomers (cDNA1-N = cDNA2-N
= 4 mM; cCu2+ = 16 mM; cRNA = 1 mM; 16 h at 37 uC).
cRNA = cCu2+ = 1 mM; 37 uC, 20 h). Calcd. masses: DNA1-Y 2726, DNA2-
Y 3022, 59-r(UGUGGAAGA)cPhos 2981, r(GUGGUUGA)-39 2566.
This journal is ß The Royal Society of Chemistry 2008
Chem. Commun., 2008, 762–764 | 763

View Article Online
mechanism of action and efficiency of even well known systems.
Additionally, despite the small library tested, we have identified a
system that is comparable to the best artificial nuclease so far
reported.
We greatly appreciate funding from The Danish National
Research Foundation. Excellent technical assistance from Lykke
H. Hansen is sincerely acknowledged. This work was supported by
The Sixth Framework Programme Marie Curie Host Fellowships
for Early Stage Research Training, under contract number MEST-
CT-2004-504018.
Notes and references
1 T. Niittyma¨ki and H. Lonnberg, Org. Biomol. Chem., 2006, 4, 15–25.
2 B. N. Trawick, A. T. Daniher and J. K. Bashkin, Chem. Rev., 1998, 98,
939–960.
3 D. S. Sigman, A. Mazumder and D. M. Perrin, Chem. Rev., 1993, 93,
2295–2316.
Fig. 6 Time dependent cleavage at different artificial nuclease : RNA
ratios (cDNA1-Y = cDNA2-Y = cCu2+; cRNA = 1 mM; 37 uC).
4 D. S. Sigman, T. W. Bruice, A. Mazumder and C. L. Sutton, Acc.
Chem. Res., 1993, 26, 98–104.
5 N. H. Williams, B. Takasaki, M. Wall and J. Chin, Acc. Chem. Res.,
1998, 32, 485–493.
6 F. Mancin, P. Scrimin, P. Tecilla and U. Tonellato, Chem. Commun.,
2005, 2540–2548.
7 S. Sakamoto, T. Tamura, T. Furukawa, Y. Komatsu, E. Ohtsuka,
M. Kitamura and H. Inoue, Nucleic Acids Res., 2003, 31, 1416–1425.
half-lifetime was found to be about two hours which corresponds
to a relatively efficient cleavage compared to other reported
artificial nucleases.1,2,7 Additionally it is clear that even at
0.5 equivalent of the DNA strands relative to RNA, over 80%
of the target was degraded after 20 h, demonstrating multiple
turnover.
˚
8 H. Astro¨m and R. Stro¨mberg, Org. Biomol. Chem., 2002, 2,
In conclusion, we have introduced a novel combinatorial
principle for identifying efficient dinuclear artificial ribonuclease
constructs by parallel screening. The principle offers the possibility
of being further developed into a high-throughput approach by
including larger libraries of nuclease constituents. In our opinion
this would represent the ultimate strategy, since as we have also
shown in this study, it is very difficult to predict the exact
1901–1907.
9 H. Aurup, T. Tuschl, F. Benseler, J. Ludwig and F. Eckstein, Nucleic
Acids Res., 1994, 22, 20–24.
10 S. K. Singh, R. Kumar and J. Wengel, J. Org. Chem., 1998, 63,
10035–10039.
11 M. Petersen and J. Wengel, Trends Biotechnol., 2003, 21, 74–81.
12 N. Kalra, B. R. Babu, V. S. Parmar and J. Wengel, Org. Biomol. Chem.,
2004, 2, 2885–2887.
764 | Chem. Commun., 2008, 762–764
This journal is ß The Royal Society of Chemistry 2008