Sequence-specific RNA cleavage by PNA conjugates of the metal-free artificial ribonuclease tris(2-aminobenzimidazole)

Article abstract of DOI:10.3762/bjoc.11.55

Tris(2-aminobenzimidazole) conjugates with antisense oligonucleotides are effective site-specific RNA cleavers. Their mechanism of action is independent of metal ions. Here we investigate conjugates with peptide nucleic acids (PNA). RNA degradation occurs

Full text of DOI:10.3762/bjoc.11.55

Sequence-specific RNA cleavage by PNA conjugates of the
metal-free artificial ribonuclease tris(2-aminobenzimidazole)
Friederike Danneberg1, Alice Ghidini2, Plamena Dogandzhiyski1, Elisabeth Kalden1,
Roger Strömberg2 and Michael W. Göbel*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2015, 11, 493–498.
1Institute for Organic Chemistry and Chemical Biology, Goethe
Universität Frankfurt, Max-von-Laue-Straße 7, D-60438 Frankfurt am
Main, Germany and 2Karolinska Institutet, Department of Biosciences
and Nutrition, Novum, SE-141 83 Huddinge, Sweden
doi:10.3762/bjoc.11.55
Received: 29 January 2015
Accepted: 02 April 2015
Published: 16 April 2015
Email:
Michael W. Göbel* - m.goebel@chemie.uni-frankfurt.de
This article is part of the Thematic Series "Nucleic acid chemistry".
Guest Editor: H.-A. Wagenknecht
* Corresponding author
Keywords:
© 2015 Danneberg et al; licensee Beilstein-Institut.
License and terms: see end of document.
antisense; fluorescence correlation spectroscopy; guanidine; miRNA
20a; peptide nucleic acids
Abstract
Tris(2-aminobenzimidazole) conjugates with antisense oligonucleotides are effective site-specific RNA cleavers. Their mechanism
of action is independent of metal ions. Here we investigate conjugates with peptide nucleic acids (PNA). RNA degradation occurs
with similar rates and substrate specificities as in experiments with DNA conjugates we performed earlier. Although aggregation
phenomena are observed in some cases, proper substrate recognition is not compromised. While our previous synthesis of
2-aminobenzimidazoles required an HgO induced cyclization step, a mercury free variant is described herein.
Introduction
Sequence specific artificial ribonucleases are an attractive oligonucleotide analogues such as peptide nucleic acids (PNA)
research target for several reasons. On the one hand, they can [4]. The benefits of oligonucleotide analogues are higher
improve our mechanistic understanding of natural ribo- affinity towards RNA [5] as well as improved resistance against
nucleases and of phosphodiester reactivity in general. On the biodegradation [6]. Furthermore, PNA can be applied to block
other hand, a wide range of practical applications is conceiv- miRNA functions in cell culture experiments [7].
able for such RNA cleavers ranging from tools in molecular
biology to chemotherapeutics targeting mRNAs or miRNAs. PNA conjugates of both metal-containing and metal-free RNA
Although most site-selective artificial ribonucleases consist of a cleavers have been effectively used as artificial nucleases
catalytic unit (both with and without metal ions) attached to a towards different substrates [8-11]. Most notable is a copper-
DNA oligonucleotide or 2’-O-methyloligoribonucleotide based PNAzyme which cleaves RNA site-specifically with sub-
complementary to the targeted RNA [1-3], possible future in strate half-lifes as low as 30 minutes [12]. A PNA-10mer at-
cell applications suggest the conjugation of RNA cleavers to tached to diethylenetriamine showed a higher cleavage activity
493

Beilstein J. Org. Chem. 2015, 11, 493–498.
[13] when compared to the corresponding 22mer DNA conju-
gate [14]. We therefore decided to investigate the activity of
tris(2-aminobenzimidazole) [15] – a metal-free ribonuclease
developed by us formerly – as a conjugate with PNA. Previous
results for the corresponding DNA conjugates had shown
sequence specific RNA cleavage with substrate half-lives in the
range of 12 to 17 h [16]. Tris(2-aminobenzimidazole) 7 can be
obtained in a six-step reaction sequence starting from tris(2-
aminoethyl)amine (1) [15]. However, the use of toxic
mercury(II) oxide for the formation of the benzimidazole
moieties (Scheme 1) is a disadvantage of this synthesis espe-
cially with regard to future biological applications of the final
product. Furthermore, conditions for these reactions are rather
harsh and the yields may differ considerably. Here we report a
new and mercury-free synthesis of tris(2-aminobenzimidazole)
and its conjugation to PNA oligomers. In addition, the results of
cleaving experiments with three different RNA substrates are
presented.
Scheme 1: Formation of the 2-aminobenzimidazole moiety.
N,N’-diisopropylcarbodiimide (DIC) [17]. Another method
developed by Lipton et al. uses Mukaiyama's reagent (2-chloro-
1-methylpyridinium iodide) to prepare guanidines in high yields
[18]. The reaction proceeds at ambient temperature. In contrast
to ureas and thioureas formed as byproducts when using
carbodiimides, the N-methylpyridine-2(1H)-thione resulting
from Mukaiyama's reagent is soluble and can be easily removed
by chromatography. As we were searching for a heavy metal-
free approach towards tris(2-aminobenzimidazole), we adopted
Lipton's method to the synthesis of aminobenzimidazoles.
Thiourea 3 was prepared as described before [15], starting with
the Boc-protection of tris(2-aminoethyl)amine (1) followed by
Results and Discussion
Alternative reagents to convert thioureas into guanidines the stepwise reaction of the free NH2 group of product 2 with
include metal salts like CuCl2 or HgCl2 and carbodiimides such thiocarbonyldiimidazole and methyl 3,4-diaminobenzoate
as 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI) or (Scheme 2). The cyclisation to form the first benzimidazole unit
Scheme 2: Synthesis of tris(2-aminobenzimidazole). Conditions: a: Boc-ON, THF, 0 °C to rt, 46 h, 45%; b: 1) 1,1‘-thiocarbonyldiimidazole, imidazole,
MeCN, 0 °C to rt, 1 h; 2) methyl 3,4-diaminobenzoate, MeCN, 50 °C, 4 h, rt, overnight, 79%; c: Mukaiyama’s reagent, NEt3, DMF, 20 h, rt, 88%;
d: 1) SOCl2, MeOH, 0 °C to rt, 3 h; 2) 2-nitrophenylisothiocyanate, NEt3, EtOH, 0 °C to rt, overnight, 90%; e: 45 bar H2, Pd/C (10%), MeOH sat. with
NH3, 60 °C, 5 h, rt, overnight, 37%; f: Mukaiyama’s reagent, NEt3, DMF, 22 h, rt, 56%. Boc-ON: 2-(tert-butoxycarbonyloxyimino)-2-phenylacetonitrile.
494

Beilstein J. Org. Chem. 2015, 11, 493–498.
was then achieved by reacting thiourea 3 with Mukaiyama’s
reagent in the presence of NEt3 at rt yielding 88% of 4.
Although the yield is slightly lower compared to the cyclisation
with HgO (97% yield), the mild and metal-free reaction condi-
tions are a big advantage of this new method. For the subse-
quent two conversions, again the known procedures were used.
Thiourea 6 was prepared by deprotection of 4 with SOCl2 in
MeOH and addition of 2-nitrophenylisothiocyanate followed by
reduction of the nitro groups with H2. In the last step of the syn-
thesis, again Mukaiyama’s reagent was used for the formation
of the two benzimidazole moieties. Stirring at rt for 22 h in the
presence of NEt3 yielded 56% of tris(2-aminobenzimidazole) 7
after HPLC purification. Ester hydrolysis with aq HCl led to the
free carboxylic acid 8, which after removal of the solvent was
used for conjugation to PNA without further purification
(Scheme 3).
Conjugation of tris(2-aminobenzimidazole) 8 was performed
with fully protected PNA-oligomers still bound to the solid
support. To increase the solubility of the final product, 10mers
were attached to one, 15mers to two lysine units placed at the
C-terminus. For conjugation, the Fmoc-protected 6-amino-
hexanoic acid 9 was bound to the PNA at the terminal amino
group as a linker. After deprotection of the linker’s amino func-
tion, tris(2-aminobenzimidazole) was attached. In both steps,
DIC and HOBt were used as reagents. Coupling of the cleaver
proceeded in high yields with no unconjugated PNA detectable
in the MALDI mass spectra of the crude products. All conju-
gates were purified by RP-HPLC and isolated as trifluoro-
Scheme 3: Synthesis of PNA conjugates. Conditions: a: 1) 9, HOBt,
DIC, DMF, rt, 24 h; 2) piperidine, DMF, rt, 30 min; b: 1) 8, HOBt, DIC,
NEt3, DMF, rt, 72 h; 2) TFA/H2O/triisopropysilane (95:2.5:2.5), rt, 3 h.
acetate salts. Starting from 20–30 mg of Rink amide MBHA Cleavage experiments were run with Cy5-labeled RNA, the
resin (0.66 mmol/g) even manual synthesis routinely yielded fluorescent dye permitting the detection and quantification of
milligram amounts of the purified final product. This compares fragments by gel electrophoresis in a DNA sequencer
favorably to the yield of DNA conjugates we have studied (Figure 1). Substrates 15 and 16 have been used for comparison
earlier [16].
with our previous studies [16]. The third substrate 17 has the
Figure 1: Sequences of PNA conjugates 10–14 and oligonucleotides 15–20. Lysines are attached to the C-terminus, linker and cleaver to the
N-terminus. The parts with base complementarity to 15mer PNAs, i.e., the binding sites, are underlined. When 10mer PNAs bind to substrates 15 and
16, the first five ribonucleotides at the 5’ end will remain single stranded.
495

Beilstein J. Org. Chem. 2015, 11, 493–498.
sequence of miRNA 20a, a member of the oncogenic miRNA
17–92 cluster [19] that is a promising target for site-specific
RNA cleavers. All RNA parts of substrates 15–17 are
embedded into a stretch of nonreacting deoxythymidine
residues to improve the separation and resolution of substrate
and fragment peaks in the sequencer.
Incubation of RNA 15 with conjugate 12, which has the same
15mer sequence as a previously tested DNA conjugate [16], led
to 8 dye labeled fragments, all located in the non-hybridized
part of the substrate. Most prominent is cleavage after G(18)
(Figure 2, lane b; note the slight shift and duplication of the
substrate peak compared to the hydrolysis ladder shown in lane
f). The product distribution is considerably broader compared to
degradation of RNA 15 by the corresponding DNA conjugate.
Similar results were obtained by reaction of substrate 17 with
15mer-conjugate 14 (lane e). Here, A(18) and G(19) are the
main cleaving sites. Unexpectedly, no distinct cleavage pattern
could be observed with the 15mer-conjugate 13 and its cognate
RNA 16. Instead, the electropherogram showed a shift of
several signals including the substrate peak (lane d, compare to
hydrolysis ladder of 16 in lane g), making it difficult to assign
Figure 2: Cleavage of RNA by their corresponding PNA conjugates
(150 nM substrate, 750 nM conjugate, 50 mM Tris-HCl, pH 8, 37 °C,
20 h). Lane a: conjugate 10 and substrate 15. Lane b: conjugate 12
and substrate 15. Lane c: conjugate 11 and substrate 16. Lane d:
conjugate 13 and substrate 16. Lane e: conjugate 14 and substrate 17.
Lanes f, g and h: hydrolysis patterns of 15, 16 and 17 (Na2CO3).
the signals to individual cleavage fragments. However, the pos- cleavage occurs in a nonspecific way in nearly all positions not
ition of the signals relative to the substrate peak suggests that protected by hybridization with PNA (lane c). Though site
cleavage occurred only in the non-hybridized part of the RNA. specificity is low in this case, conjugate 13 is nevertheless sub-
We explain this effect by incomplete PNA–RNA strand sep- strate specific and does not degrade RNAs 15 or 17 under iden-
aration during electrophoresis due to stronger hybridization tical conditions (Supporting Information File 1). The relatively
resulting from the G/C-rich sequence of the conjugate (10 G/C- low site specificity, as discussed above, may be related to the
base pairs in contrast to 7 and 4 for conjugates 12 and 14, res- pyrimidine rich sequence of RNA 16.
pectively). The fact that incubation of RNA 15 with PNA 12
also leads to a slight, but much weaker shift of signals (lane b), To examine whether PNA conjugates 10–14 would form molec-
supports this assumption. The first four nucleotides in the single ular aggregates with noncognate oligonucleotides, the diffusion
stranded part of RNA 16 where cleavage takes place have only time of Cy5-labeled DNA 18 (19 nM, diluted with 131 nM
pyrimidine bases, while RNA 15 and 17 have mainly purine unlabeled DNA 19) in absence and presence of different PNA
bases in the corresponding regions. If the tris(2-aminobenzimi- conjugates (750 nM) was studied by fluorescence correlation
dazole) cleaver interacts with the purine bases (e.g., by spectroscopy (FCS) [15,20]. No significant change in diffusion
stacking) it is not impossible that this could also be related to time was observed upon addition of 10mer conjugate 10 to
the apparent lower rate and lower selectivity in cleavage of noncognate DNAs 18/19, indicating that no aggregates were
RNA 16 with both 11 and 13. Consistent with that idea, we formed. In contrast, addition of 10mer conjugate 11 or of 15mer
previously had observed similar effects when RNAs 15 and 16 conjugates 12, 13, and 14 to DNAs 18/19 led to considerable
were cleaved by DNA-catalyst conjugates analogous to 12 and changes in diffusion times, which was most evident for PNAs
13 [16].
11 and 13. This clearly shows that molecular aggregates of
oligonucleotides and PNA conjugates are formed in all of these
As it is known that PNA binds more tightly to RNA than does cases [21]. Although such effects were absent in FCS experi-
DNA [5], 10mer-conjugates 10 and 11 complementary to ments with analogous DNA conjugates [16], this result is not
substrates 15 and 16 were also tested in cleavage experiments. surprising as PNA, with its uncharged backbone, is more likely
Compared to the 15mers, the first 5 monomers were omitted in to form aggregates. In the presence of 10% of DMSO, no
order to keep the cleavage site unchanged. Results for the incu- change in diffusion times was observed for conjugates 12 and
bation of 15 with 10 were as expected (lane a). In contrast to 14. Conjugates 11 and 13 still caused an effect but less
incubation of substrate 16 with 15mer-conjugate 13, no signal pronounced than in the absence of DMSO, indicating that
shift is seen with the 10mer-conjugate 11. However, the DMSO prevents aggregation partially. We also investigated
496

Beilstein J. Org. Chem. 2015, 11, 493–498.
RNA 16 by FCS (19 nM 16, diluted with 131 nM of the unla-
beled analog 20). Upon addition of 10mer PNA 11, a large
increase of diffusion times was observed that can only be
explained by aggregation.
Aggregation of PNA conjugates with complementary RNA but
even with noncognate oligonucleotides [21] might cause a loss
of sequence specificity. It was of critical importance, therefore,
to conduct cross reaction experiments of PNA conjugates 10–14
in all possible combinations with substrates 15–17. Exem-
plarily, Figure 3 shows tests of conjugates 12 and 14 with
RNAs 15 and 17. Consistent data was obtained for all other
possible combinations of conjugates and substrates (Supporting
Information File 1), which proves that cleavage in the exam-
ined concentration range fully depends on Watson–Crick base
pairing.
Figure 4: Cleavage of RNA substrates 15, 16, and 17 by their
matching conjugates as a function of conjugate concentration (150 nM
substrate, 62.5–2000 nM conjugate, 50 mM Tris-HCl, pH 8, 37 °C,
20 h). Data points are connected by lines for the sake of clarity.
For the reaction of substrate 15 with conjugate 12, cleavage
kinetics were studied in detail. Figure 5 shows RNA decay
under saturation conditions as a function of time. While in the
absence of a cleaver the substrate proved to be quite stable for
several days, it was cleaved significantly by conjugate 12 within
a few hours. Almost complete degradation was achieved after
60 h. As seen before [16], a small percentage of RNA remained
intact even after one week. This could be due to structural
imperfections in the synthetic substrate RNA preventing the
PNA conjugate from hybridizing with its target molecule. The
best fit of data to a first order rate law is shown as a solid curve:
k1 = 0.062 h−1, t1/2 = 11.2 h. This number comes close to the
Figure 3: Substrate specificity of conjugates 12 and 14 (150 nM sub-
strate, 750 nM conjugate, 50 mM Tris-HCl, pH 8, 37 °C, 20 h). Lane a:
conjugate 12 and substrate 15. Lane b: conjugate 14 and substrate 15.
Lane c: conjugate 14 and substrate 17. Lane d: conjugate 12 and sub-
strate 17. Lanes e and f: hydrolysis patterns of 15 and 17 (Na2CO3).
Cleavage experiments with varying amounts of conjugates
(Figure 4) show that – at concentrations below 1.5 µM – RNA
cleavage by PNAs 10–12 and 14 obeys saturation kinetics. Full
activity is reached at concentrations of 500–750 nM. In com-
parison to the shortened analog 10, 15mer conjugate 12 shows
higher cleavage yields of RNA 15 at concentrations from 62.5
to 1000 nM. Larger amounts of PNA, however, decrease the
effectiveness of RNA degradation. This drop is quite consider-
able for the 15mer conjugates, whereas it is less significant for
the 10mer analogs. The decreasing cleavage yields at high PNA
concentrations are likely to result from increased aggregation,
which might prevent the cleaver from interacting with the RNA
backbone.
Figure 5: Cleavage kinetics of 15 in the presence and absence of
conjugate 12. Conditions: 150 nM substrate, 0 or 750 nM conjugate,
50 mM Tris-HCl, pH 8, 100 mM NaCl, 37 °C. The solid curves are
calculated assuming first order kinetics.
497

Beilstein J. Org. Chem. 2015, 11, 493–498.
7. Torres, A. G.; Fabani, M. M.; Vigorito, E.; Williams, D.; Al-Obaidi, N.;
Wojciechowski, F.; Hudson, R. H. E.; Seitz, O.; Gait, M. J.
Nucleic Acids Res. 2012, 40, 2152–2167. doi:10.1093/nar/gkr885
8. Whitney, A.; Gavory, G.; Balasubramanian, S. Chem. Commun. 2003,
36–37. doi:10.1039/b210135a
half-life of 15 when cleaved by the analogous DNA conjugate
(t1/2 = 12.4 h) [16].
Conclusion
We have shown that PNA conjugates of tris(2-aminobenzimida-
zole) are potent hydrolytic cleavers of RNA, exhibiting similar
efficiency and substrate specificity as the corresponding DNA
conjugates tested previously. Our optimized synthesis of tris(2-
aminobenzimidazole) provides a new way to obtain the catalyst
avoiding mercury salts with all their disadvantages (toxicity,
harsh reaction conditions, variable yields). Conjugation to PNA
oligomers can be readily achieved in high yields, providing an
easy way to synthesize sequence specific metal-free artificial
nucleases for a wide range of RNA substrates. Due to the ten-
dency of PNA conjugates to form aggregates, a phenomenon
not seen with their DNA analogs, optimization of the PNA
oligomer length is necessary, especially for strands rich in G/C
base pairs. Reduced activity of long PNA conjugates caused by
aggregation can be a minor disadvantage in comparison to DNA
conjugates. However, the higher affinity of PNA towards RNA
allows using shorter PNA oligomers. In addition, the stability of
PNA against biodegradation is another advantage especially
with respect to possible in cell applications.
9. Riguet, E.; Tripathi, S.; Chaubey, B.; Désiré, J.; Pandey, V. N.;
Décout, J.-L. J. Med. Chem. 2004, 47, 4806–4809.
doi:10.1021/jm049642d
10.Gaglione, M.; Milano, G.; Chambery, A.; Moggio, L.; Romanelli, A.;
Messere, A. Mol. BioSyst. 2011, 7, 2490–2499.
doi:10.1039/c1mb05131h
11.Murtola, M.; Strömberg, R. Org. Biomol. Chem. 2008, 6, 3837–3842.
doi:10.1039/b810106j
12.Murtola, M.; Wenska, M.; Strömberg, R. J. Am. Chem. Soc. 2010, 132,
8984–8990. doi:10.1021/ja1008739
13.Petersen, L.; de Koning, M. C.; van Kuik-Romeijn, P.; Weterings, J.;
Pol, C. J.; Platenburg, G.; Overhand, M.; van der Marel, G. A.;
van Boom, J. H. Bioconjugate Chem. 2004, 15, 576–582.
doi:10.1021/bc034219p
14.Komiyama, M.; Inokawa, T.; Yoshinari, K.
J. Chem. Soc., Chem. Commun. 1995, 77–78.
doi:10.1039/c39950000077
15.Scheffer, U.; Strick, A.; Ludwig, V.; Peter, S.; Kalden, E.; Göbel, M. W.
J. Am. Chem. Soc. 2005, 127, 2211–2217. doi:10.1021/ja0443934
16.Gnaccarini, C.; Peter, S.; Scheffer, U.; Vonhoff, S.; Klussmann, S.;
Göbel, M. W. J. Am. Chem. Soc. 2006, 128, 8063–8067.
doi:10.1021/ja061036f
17.Katritzky, A. R.; Rogovoy, B. V. ARKIVOC 2005, No. iv, 49–87.
doi:10.3998/ark.5550190.0006.406
18.Yong, Y. F.; Kowalski, J. A.; Lipton, M. A. J. Org. Chem. 1997, 62,
1540–1542. doi:10.1021/jo962196k
Supporting Information
19.Lagos-Quintana, M.; Rauhut, R.; Lendeckel, W.; Tuschl, T. Science
2001, 294, 853–858. doi:10.1126/science.1064921
20.Müller, J. D.; Chen, Y.; Gratton, E. Methods Enzymol. 2003, 361,
69–92. doi:10.1016/S0076-6879(03)61006-2
Supporting Information File 1
Experimental procedures and characterization data.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-11-55-S1.pdf]
21.Sahu, B.; Sacui, I.; Rapireddy, S.; Zanotti, K. J.; Bahal, R.;
Armitage, B. A.; Ly, D. H. J. Org. Chem. 2011, 76, 5614–5627.
doi:10.1021/jo200482d
Acknowledgements
Financial support by the European 7th framework program
(238679, PhosChemRec), the Swedish Research Council and by
the state of Hesse (LOEWE, SynChemBio) is gratefully
acknowledged.
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
References
1. Lönnberg, H. Org. Biomol. Chem. 2011, 9, 1687–1703.
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
doi:10.1039/c0ob00486c
2. Åström, H.; Williams, N. H.; Strömberg, R. Org. Biomol. Chem. 2003, 1,
1461–1465. doi:10.1039/b212216b
3. Åström, H.; Strömberg, R. Org. Biomol. Chem. 2004, 2, 1901–1907.
doi:10.1039/b403652b
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
4. Nielsen, P. E.; Egholm, M. Curr. Issues Mol. Biol. 1999, 1, 89–104.
5. Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freier, S. M.;
Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden, B.; Nielsen, P. E. Nature
1993, 365, 566–568. doi:10.1038/365566a0
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
6. Demidov, V. V.; Potaman, V. N.; Frank-Kamenetskii, M.; Egholm, M.;
Buchardt, O.; Sönnichsen, S. H.; Nielsen, P. E. Biochem. Pharmacol.
1994, 48, 1310–1313. doi:10.1016/0006-2952(94)90171-6
doi:10.3762/bjoc.11.55
498