A highly acidic acridine for efficient site-selective activation of RNA leading to an eminent ribozyme mimic

Article abstract of DOI:10.1016/S0040-4039(02)02017-8

9-Amino-2-methoxy-6-nitroacridine (1a) is conjugated with oligonucleotide for site-selective RNA hydrolysis. When this conjugate forms a duplex with complementary RNA, the phosphodiester linkage of the RNA in front of 1a is activated and selectively hydro

Full text of DOI:10.1016/S0040-4039(02)02017-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 8249–8252
A highly acidic acridine for efficient site-selective activation of
RNA leading to an eminent ribozyme mimic
Akinori Kuzuya, Kenzo Machida and Makoto Komiyama*
Research Center for Advanced Science and Technology, The University of Tokyo, 4-6-1 Komaba, Meguro, Tokyo 153-8904,
Japan
Received 30 August 2002; revised 14 September 2002; accepted 17 September 2002
Abstract—9-Amino-2-methoxy-6-nitroacridine (1a) is conjugated with oligonucleotide for site-selective RNA hydrolysis. When
this conjugate forms a duplex with complementary RNA, the phosphodiester linkage of the RNA in front of 1a is activated and
selectively hydrolyzed by Lu(III) ion. Covalent fixation of the metal ion to sequence-recognizing moiety is unnecessary. The
site-selective hydrolysis by this conjugate is 2.2 times as fast as that by the oligonucleotide bearing 9-amino-6-chloro-2-methoxy-
acridine (1b), which hitherto has been the most active for the RNA activation. The acridine derivative 1a is more acidic than 1b,
and thus is more effective as acid catalyst for the RNA hydrolysis. © 2002 Elsevier Science Ltd. All rights reserved.
Preparation of site-selective artificial ribonucleases has
been a subject of growing interest. In almost all those
hitherto reported, a catalyst for RNA hydrolysis was
tethered to oligonucleotide as sequence-recognizing
moiety.1 Recently, we have developed an entirely new
strategy, which involves no covalent fixation of catalyst
to oligonucleotide.2,3 The target phosphodiester linkage
in substrate RNA is activated through non-covalent
interactions with oligonucleotide bearing an acridine,
and differentiated from the other linkages in terms of
reactivity. Thus, this linkage is selectively hydrolyzed
even when catalysts (e.g. lanthanide(III) ions) are not
fixed anywhere. The site-selective scission is highly
efficient, mainly because the catalysts are free from
complexation with strong ligands. Consistently, rather
poor active catalysts such as Zn(II) ion and Mn(II) ion
also showed sufficient catalysis. Previous kinetic study
indicated that these non-covalent site-selective hydroly-
ses involve the acid catalysis by the acridine in its
protonated form.3 Among all the acridine derivatives
investigated, 9-amino-6-chloro-2-methoxyacridine (1b),
which has the smallest pKa (=8.5),4 was the most
active. According to this result, still faster site-selective
RNA hydrolysis should be possible if more acidic
acridine derivative is tethered to oligonucleotide. Here
we report that oligonucleotide bearing 9-amino-2-
methoxy-6-nitroacridine (1a: pKa=7.3)4 shows enor-
mously fast site-selective RNA hydrolysis. This result
confirms the previously proposed acid catalysis by the
acridines, providing a strong basis for the molecular
design towards still more efficient RNA processing.
Oligonucleotides used in this study are presented in Fig.
1. The acridine–DNA conjugate (DNA1–1a) bears 9-
amino-2-methoxy-6-nitroacridine (1a) at the 5%-end of
18-mer oligonucleotide which is complementary to the
5%-side of the 36-mer substrate RNA. The unmodified
17-mer oligonucleotide (DNA2) is complementary to the
remaining portion of the RNA. When the RNA is
hybridized to both of DNA1–1a and DNA2, only U19
remains unpaired as the target site. Another acridine–
DNA conjugate (DNA1–1b) involves 9-amino-6-chloro-
Keywords: acridine; oligonucleotides; site-selective scission; lan-
thanide; acid catalysis.
* Corresponding author. Tel.: +81-3-5452-5200; fax: +81-3-5452-5209;
e-mail: komiyama@mkomi.rcast.u-tokyo.ac.jp
Figure 1. Oligonucleotides used in this study.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02017-8

8250
A. Kuzuya et al. / Tetrahedron Letters 43 (2002) 8249–8252
Scheme 1. Reagents and conditions for the synthesis of the phosphoramidite monomer 5 bearing 1a: (i) Cu, CuO, K2CO3, H2O,
reflux, 12 h; (ii) POCl3, reflux, 2 h; (iii) p-chlorophenol, 115°C, 12 h; (iv) 6-amino-1-hexanol, DMF, 70°C, 12 h; (v)
(iPr2N)2POEtCN, 1H-tetrazole, DMF, rt, 1 h.
2-methoxyacridine (1b) in place of 1a, and, as described
above, has been the most effective RNA activator until
the present study.
sponding band in lane 1 for the random scission by
Lu(III), the site-selective activation of the target phos-
phodiester linkage by the DNA1–1a/DNA2 combination
is evident. Another acridine–DNA conjugate DNA1–1b,
which bears 9-amino-6-chloro-2-methoxyacridine (1b),
also induces site-selective RNA hydrolysis (lane 3). It is
noteworthy that the site-selective hydrolysis by the
DNA1–1a/DNA2/Lu(III) system is significantly faster
than that by the DNA1–1b/DNA2/Lu(III) system (Fig.
3). Pseudo-first-order rate constants are 0.22 and 0.10
All the oligonucleotides were synthesized using stan-
dard phosphoramidite chemistry on an automated syn-
thesizer. The synthetic route for the phosphoramidite
monomer bearing 1a (5) is shown in Scheme 1. Com-
mercially available 2-chloro-4-nitrobenzoic acid was
coupled with p-methoxyaniline to give 2, which was
cyclized by using phosphoryl chloride. The resultant 3
was heated in p-chlorophenol, and then with 6-amino-
1-hexanol in DMF to give the linker-bearing acridine
4,5 which was quantitatively converted to the phospho-
ramidite monomer 5. Attachment of 5 to oligonucle-
otide was performed in the final cycle of the DNA
synthesis with an extended coupling time of 10 min.
After the synthesis, DNA1–1a was detached from the
support and deprotected with 0.4 M methanolic NaOH
(4:1 MeOH/H2O) at room temperature for 16 h, and
the mixture was desalted. DNA1–1a was purified by a
reversed-phase HPLC equipped with an RP-C18
column, and characterized by MALDI-TOF MS analy-
ses (m/z, 5862.9; calcd [M−H]−, 5862.0). In addition,
the acridine unit 4 (m/z, 370) was detected by ESI MS,
when DNA1–1a was digested with snake venom phos-
phodiesterase and alkaline phosphatase. The digest was
further analyzed by the HPLC, providing the expected
base composition (A:T:G:C=4:5:3:6).
The site-selective RNA hydrolyses using the modified
oligonucleotides as RNA activators were performed
with 100 mM LuCl3 at pH 8.0 and 37°C in the presence
of 200 mM NaCl.6 After 2 h, the reactions were
quenched by 10 mM EDTA-2Na and analyzed by
denaturing PAGE. The scission efficiency was evalu-
ated by a Fuji Film FLA-3000G fluorescent imaging
analyzer. As shown in lane 4 of Fig. 2, the substrate
RNA is site-selectively and efficiently hydrolyzed by
DNA1–1a/DNA2/Lu(III) system. The hydrolysis occurs
mainly at the 5%-side of U19 as the target site (the minor
scission is perceived at its 3%-side).7 By comparing the
intensity of this major band with that of the corre-
Figure 2. Site-selective scission of the RNA by the combina-
tions of modified or unmodified DNA1, DNA2, and Lu(III) at
37°C and pH 8.0 in 2 h; [RNA]0=1, [DNA]=10 and
[Lu(III)]=100 mM. Lane 1, treatment with Lu(III) alone; lane
2, DNA1/DNA2/Lu(III); lane 3, DNA1–1b/DNA2/Lu(III); lane
4, DNA1–1a/DNA2/Lu(III). R, no treatment of the RNA; H,
alkaline hydrolysis; T1, RNase T1 digestion; C, control in
buffer solution.

A. Kuzuya et al. / Tetrahedron Letters 43 (2002) 8249–8252
8251
Table 1. Melting temperatures of the duplex between the
a
RNA and modified or unmodified DNA1
Tm (°C)
DTm (°C)b
DNA1–1a
DNA1–1b
DNA1
68.0
69.1
63.1
4.9
6.0
–
a Conditions: [RNA]=1.0 mM, [modified or unmodified DNA1]=1.0
mM, [NaCl]=200 mM, and [Tris–HCl]=10 mM (pH 8.0).
b Differences of Tm’s from the value for the unmodified DNA1.
DNA1–1b, simply because its intrinsic activity as acid
catalyst is greater as expected from the Brønsted rule.
As shown in Table 1, the melting temperatures (Tm’s)
of the RNA/DNA1–1a and the RNA/DNA1–1b het-
eroduplexes are higher than that of the RNA/DNA1
duplex, and the duplex-stabilizing activity of the
acridine in DNA1–1a is smaller than that for DNA1–1b
(DTm=4.9 and 6.0°C, respectively). It is unlikely that
DNA1–1a is more active for the RNA hydrolysis
because 1a interacts with the RNA more strongly than
does 1b.
Figure 3. Time-course of the site-selective RNA hydrolysis at
U19. Filled circles, hydrolysis with DNA1–1a/DNA2/Lu(III);
open circles, hydrolysis with DNA1–1b/DNA2/Lu(III).
h−1, respectively. Site-selective RNA activation by 1a,
synthesized according to Scheme 1, sufficiently sur-
passes that by commercially available 1b.
In conclusion, highly efficient RNA activator has been
This notably high activity of the DNA1–1a/DNA2/
Lu(III) system is attributable to the high acidity of 1a,
as substantiated by the following results. The pKa val-
ues of the acridine residues in DNA1–1a and DNA1–1b
were directly evaluated by using their absorbances (Fig.
4). Curve fitting of the experimental points gives the
pKa values of 8.8 0.1 for DNA1–1a8 and 10.5 0.1 for
DNA1–1b. Upon the attachment of acridines to
oligonucleotides, their pKa values are elevated by 1.5–
2.0 units, compared to that of the corresponding free
acridine. These pKa values show that almost 100% of 1b
in DNA1–1b should be protonated under the reaction
conditions (pH 8.0) and potent for the acid catalysis.
On the other hand, the fraction of protonated form for
1a in DNA1–1a is smaller (86%). Apparently, DNA1–1a
is more active for the present RNA hydrolysis than
synthesized
by
attaching
9-amino-2-methoxy-6-
nitroacridine to oligonucleotide. The site-selective RNA
hydrolysis by this conjugate is among the fastest ever
reported. The present finding should serve as important
information for the design of useful tools for biotech-
nology and molecular biology. Studies of their practical
applications are currently underway in our laboratory.
Acknowledgements
This work was partially supported by Bio-oriented
Technology Research Advancement Institution.
A
Grant-in-Aid for Scientific Research from the Ministry
of Education, Science, Sports, Culture and Technology,
Japan, is also acknowledged.
References
1. Reviews: (a) Komiyama, M.; Sumaoka, J.; Kuzuya, A.;
Yamamoto, Y. Methods Enzymol. 2001, 341, 455–468; (b)
Cowan, J. A. Curr. Opin. Chem. Biol. 2001, 5, 634–642; (c)
Trawick, B. N.; Daniher, A. T.; Bashkin, J. K. Chem. Rev.
1998, 98, 939–960; (d) Oivanen, M.; Kuusela, S.;
Lo¨nnberg, H. Chem. Rev. 1998, 98, 961–990; (e) Ha¨ner,
R.; Hall, J.; Pfu¨tzer, A.; Hu¨sken, D. Pure Appl. Chem.
1998, 70, 111–116.
2. Kuzuya, A.; Machida, K.; Mizoguchi, R.; Komiyama, M.
Bioconjugate Chem. 2002, 13, 365–369.
3. Kuzuya, A; Mizoguchi, R.; Morisawa, F.; Machida, K.;
Komiyama, M. J. Am. Chem. Soc. 2002, 124, 6887–6894.
4. Albert, A. The Acridines, 2nd ed.; Edward Arnold Publish-
ers: London, 1966; pp. 434–467. Note that these pKa
values are the ones for free acridine derivatives. As
described in the text, they are elevated on their attachment
to oligonucleotide due to electrostatic interactions with the
anionic phosphates.
Figure 4. The pH dependence of the absorbance of acridines.
Filled circles, DNA1–1a (at 476 nm); open circles, DNA1–1b
(at 451 nm).

8252
A. Kuzuya et al. / Tetrahedron Letters 43 (2002) 8249–8252
5. Spectral data of 4: TLC Rf=0.22, 10:1 CH2Cl2/MeOH;
1H NMR (DMSO-d6) l 8.67 (s, 1H), 8.53 (d, 1H),
7.98 (d, 1H), 7.93 (d, 1H), 7.68 (s, 1H), 7.49 (d, 1H),
4.32 (t, 1H), 3.96 (s, 3H, OMe), 3.78 (q, 2H), 1.73
(m, 2H), 1.36–1.25 (m, 6H). MS (ESI) m/z 370
[M+H]+.
6. Light exposure was carefully avoided during the hydrolysis
reaction.
7. In lane 4, the scission at the 3%-side of U19 is 10-fold less
efficient than that at its 5%-side.
8. The pKa of 1a in RNA/DNA–1a heteroduplex is almost
identical with this value.