Bicyclo[3.2.1]amide-DNA: Synthesis and base-pairing properties

Article abstract of DOI:10.1055/s-1999-3113

A new oligonucleotide analog, bicyclo[3.2.1]-amide-DNA was synthesized. A homo-adenylate sequence of this analog, efficiently forms hetero-duplexes with complementary natural DNA and RNA. The corresponding homo-thymidylate sequence, however, does not base

Full text of DOI:10.1055/s-1999-3113

LETTER
913
Bicyclo[3.2.1]amide-DNA: Synthesis and Base-Pairing Properties
Anita Egger, Christian J. Leumann
Department of Chemistry and Biochemistry, University of Bern, Freiestrasse 3, CH-3012 Bern
Fax +41(31)6313422; E-mail: leumann@ioc.unibe.ch
Received 10 February 1999
(EDI) as the coupling reagent in DMF gave the nucleoside
analog 2 (80%) that was subsequently tritylated with 4,4‘-
(dimethoxy)trityl triflate6 to give 4 in high yield. Desily-
lation of 4 with Bu4NF in THF afforded 6 (83%) that was
Abstract: A new oligonucleotide analog, bicyclo[3.2.1]-amide-
DNA was synthesized. A homo-adenylate sequence of this analog,
efficiently forms hetero-duplexes with complementary natural
DNA and RNA. The corresponding homo-thymidylate sequence,
however, does not base-pair to complementary DNA and RNA, but finally converted to the phosphoramidite 8 in 86% yield.
forms a stable duplex with the homo-adenylate oligomer of the
same backbone type. The structure of this duplex is virtually enan-
tiomeric to that of a natural B-DNA, as inferred from CD-spectros-
mixed anhydride between N6-benzoyladen-9-yl acetic
copy.
Key words: oligonucleotides, DNA recognition, RNA, antisense
The synthesis of the adenine containing nucleoside analog
3 (48%) was somewhat more cumbersome in that the
acid and isobutyl chloroformate had to be used to effect
coupling to 1. The remaining reaction sequence 3 → 5 →
7 → 9 was performed as described for the thyminyl series in
an overall yield of 22%.7
agents
DNA-analogs are of interest as potential antisense agents
in medicinal chemistry, and as tools in molecular biolo-
gy.1 Despite considerable efforts made over the recent
years, the design of DNA-analogs with taylor-made bio-
logical and chemical properties still remains a challenge.
With the aim of investigating the structural role of the
furanose unit in RNA and DNA we recently designed and
synthesized the analog bicyclo[3.2.1]-DNA2 (Figure 1).
Figure
1 Structures of DNA, bicyclo[3.2.1]-DNA and bi-
cyclo[3.2.1]amide-DNA
Scheme 1 Reaction conditions: a) (2):1-Thy-CH2COOH (1.2 eq.),
EDC (1.4 eq.), DMF, r.t., 5h, 80%; (3): 9-Bz6ade-CH2COOH (1.5
eq.), N-methyl morpholine (10 eq.), ClCOOiBu (1.5 eq.), DMF, 0°C,
1 min., then 1 (1 eq.) in DMF, r.t., 16h, 48%; b) DMT-OTf (2-4 eq.),
pyridine, 50-60°C, 3-5h, 90% (4), 68% (5); c) Bu4NF (2 eq.), THF,
r.t., 12-16h, 83% (6), 88% (7); d) ClP(OCH2CH2CN)N(iPr)2 (3 eq.),
(iPr)2NEt (6 eq.), THF, r.t., 3h, 86% (8), 75% (9).
In this analog the pentofuranose unit of DNA is replaced
by a structurally rigid backbone unit that emulates a back-
bone conformation of the B-type, to which the bases were
attached via a flexible, open-chain oxymethylene linker
unit. Bicylo[3.2.1]amide-DNA is a related analog in
which the bases are attached via an amide linker to the
otherwise identical sugar scaffold. Here we describe the
successful synthesis and the first pairing properties of bi-
cyclo[3.2.1]-amide oligomers of the bases thymine and
adenine.3
The phosphoramidite building blocks 8 and 9 were then
used for the automated solid-phase assembly of oligonu-
cleotides 10 and 11 with a DNA synthesizer. Both synthe-
ses were performed on a 1.3 mmol scale and, for reasons
of synthetic convenience, started with commercially
available thymidine-loaded controlled pore glass. The
synthesis cycle for natural deoxyoligonucleotides was
used and adapted as follows: The coupling time was ex-
tended to 6 min, and detritylation was performed utilizing
The monomeric building blocks 8 and 9 of bicyc-
lo[3.2.1]amide-DNA, suitable for solid-phase oligonucle-
otide synthesis, were synthesized starting from the
bicyclic amine 1 (Scheme 1) that was prepared earlier in
our laboratory.4 Coupling of thymin-1-yl acetic acid5 with
1 using ethyl-N,N-dimethylaminoethyl carbodiimide
Synlett 1999, S1, 913–916 ISSN 0936-5214 © Thieme Stuttgart · New York

914
A. Egger, C. J. Leumann
LETTER
10% trichloroacetic acid in dichloroethane (60 sec). While the thymine-containing sequence 10 shows no de-
Yields per coupling cycle (trityl assay) were on the aver- tectable signs of complex formation with d(A10) or
age of 95-99%. After synthesis the oligonucleotides were poly(A) by both, UV-melting analysis and non-denaturing
deprotected and detached from solid support using stan- polyacrylamide gel electrophoresis, the purine sequence
dard conditions (conc. NH3, 55-60°C, 12-16 h). The crude 11 efficiently base-pairs to d(T10) and poly(U) as well as
oligomers were purified by ion-exchange HPLC, and the to 10. The corresponding Tm-data are summarized in Ta-
purity of the isolated fractions controlled by reversed ble 2. Representative UV-melting curves are depicted in
phase HPLC. MALDI-ToF mass spectrometry confirmed Figure 3.
the constitution of the two oligomers (Table 1).
The pairing properties of 10 and 11 with complementary
DNA and RNA and with each other was investigated by
UV-melting curves, by non-denaturing polyacrylamide
electrophoresis and by CD-spectroscopy. While UV-
Figure 3 UV-melting curves (260 nm) of the sequences 10 and
melting curves of the thymine containing oligomer 10
alone indicates no self-aggregation, that of the adenine
containing oligomer 11 shows a highly cooperative tran-
sition with a Tm of 10°C (Figure 2). This transition is best
visible at 284 nm and displays negative hyperchromicity
at this wavelength. The transition is independent from pH
in the range of pH 6-8, thus ruling out the presence of an
A-A+ base-paired structure as observed in RNA at low
pH.8
11 with complementary DNA and RNA and with each other (c =
4-6 mM in 10 mM Na-cacodylate, 1M NaCl, pH 7.0
A
B
C
D
E
A
B
C
D
E
Figure 4 UV-shadowing of non-denaturing 20% polyacrylami-
de gels (1 resp. 2 nmol of single strand, 10mM MgCl2, 90 mM
tris borate, pH 8.1, 4°C). Left; A: 10, B: 10:11 = 2:1, C: 10:11 =
1:1, D: 10:11 = 1:2, E: 11. Right; A: 11, B: 11:d(T10) = 2:1, C:
11:d(T10) = 1:1, D: 11:d(T10) = 1:2, E: d(T10).
Relative to d(A10), duplex formation of 11 with d(T10) or
poly(U) is reduced by only ca. 1°C/mod. (DTm/mod). The
unnatural duplex 10:11 is of almost equal thermal stability
as the duplex 11:d(T10). The stoichiometry of complex
formation for the system 11:d(T10) and 10:11 was ad-
dressed by gel mobility assays (Figure 4). Inspection of
the gels reveals that 11 pairs to 10 forming a duplex and
not a triplex. Only in the case of the 1:1 stoichiometric
mixture (Figure 4, left lane C), all the bands arising from
the single strands have disappeared. Interestingly, no
sharp but only a broad new band with very low mobility
can be seen. This could be due to catenation of 10 and 11.
Sequence 11 also forms only a duplex with d(T10). From
Figure 4, right, lanes B-D only one new structure attribut-
Figure 2 UV melting curve of 11 at 260 nm and 284 nm
(c = 4mM in 10 mM Na-cacodylate, 1M NaCl pH 7.0).
Synlett 1999, S1, 913–916 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Bicyclo[3.2.1]amide-DNA: Synthesis and Base-Pairing Properties
915
able to a bimolecular duplex arises, and excess of the sin- phase DNA-chemistry. This analog consitutes a new,
gle strands in stoichiometric mixtures different from 1:1 hitherto unknown A-T base-pairing system, with overall
are still visible. Both results were also independently con- structural properties that are similar to enantio-DNA. This
firmed by Job-plot analysis.3
is an unexpected feature since its repetitive backbone unit
was designed to emulate a D-ribose backbone in B-confor-
mation. The distincly different association properties of
this analog compared to bicyclo[3.2.1]-DNA originates
solely from the change in the linker between the base and
the backbone unit, that is longer by one bond in bicyc-
lo[3.2.1]amide-DNA compared to bicyclo[3.2.1]-DNA
(and DNA itself). Besides PNA11 and alanyl-PNA12 the
analog presented here is a further example of a pairing-
competent DNA-analog, in which a ring system between
the backbone and the nucleobase as a structurally preorga-
nizing element is missing.
We determined the thermodynamic data of duplex forma-
tion in the system 10:11 and 11:d(T10), and for compari-
son of d(A10):d(T10), (Table 3) from 1/Tm vs. ln[c] plots.9
Interestingly, only the heteroduplex 11:d(T10) shows a
pronounced reduction in the pairing enthalpy term (DH)
relative to the natural duplex. The thermodynamic stabil-
ity (DG25°C) decreases in the order d(A10):d(T10) >
11:d(T10) > 10:11 and is well reflected by the correspond-
ing Tm-data (Table 2). The higher instability of the unnat-
ural duplex 10:11 relative to d(A10):d(T10) seems to be of
entropic and not enthalpic origin.
As is the case for natural DNA, bicyclo[3.2.1]amide-DNA
can strongly discriminate between matched and mis-
matched complementary sequences. Upon pairing of 11 to
complementary DNA with a mismatched base in the cen-
ter, a considerable drop in Tm is observed. However, mis-
match discrimination in the case of bicyclo[3.2.1]amide-
DNA is slightly less effective than in the case of natural
DNA ( Table 4).
Figure 5 CD-spectra of duplexes of 10 and 11 at 4°C (ctot
4 mM, d= 1 cm, in 10mM Na-cacodylate, 1M NaCl, pH 7.0).
=
The continuing experimental exploration of new back-
bone-modified DNA-analogs will produce more comple-
mentary base-pairing systems with different functional
properties in the future, that are not only of interest as
drugs in medicinal chemistry, but also as intelligent mo-
lecular devices in materials science and computer technol-
ogy.
Acknowledgement
Financial support from the Swiss National Science Foundation,
from Novartis AG, Basel and from the Wander-Stiftung, Bern, is
gratefully acknowledged.
We investigated the structural properties of duplexes of 11
with complementary DNA, RNA and with 10 by CD spec-
troscopy (Figure 5). The duplex 11:poly(U) shows all the
characteristics of a A-DNA conformation, and the duplex
between 11 and d(T10) is very similar in shape (wave-
length but not absolute ellipticities) to that of the natural
duplex d(A10):d(T10) (Figure 5, right). Most interestingly,
however, the spectrum of the duplex in the pure non-nat-
ural series, 10:11, is almost the mirror image of that of the
natural system d(A10):d(T10) (Figure 5, left). On the basis
of CD-evidence we assign to this duplex a left-handed,
Watson-Crick base-paired, helical structure.10
References and Notes
(1) Uhlmann, E.; Peyman, A. Chem. Rev. 1990, 90, 543. De
Mesmaeker, A.; Häner, R.; Martin, P.; Moser, H. E. Acc.
Chem. Res. 1995, 28, 366. Uhlmann, E. Chem. Zeit 1998, 32,
150.
(2) Epple, C.; Leumann, C. Chem. & Biol. 1998, 5, 209.
(3) Egger, A. Dissertation, Universität Bern, 1998
(4) Egger, A.; Hunziker, J.; Rihs, G.; Leumann, C. Helv. Chim.
Acta 1998, 81, 734.
(5) Dueholm, K. L.; Egholm, M.; Behrens, C.; Christensen, L.;
Handen, H. F.; Vulpius, T.; Petersen, K. H.; Berg, R. H.;
Nielsen, P. E.; Buchardt, O. J. Org. Chem. 1994, 89, 5767.
(6) Tarköy, M.; Bolli, M.; Leumann, C. Helv. Chim. Acta 1994,
77, 716.
In conclusion we have shown here, that bicyc-
lo[3.2.1]amide-DNA containing the bases adenine and
thymine can efficiently be prepared using standard solid-
Synlett 1999, S1, 913–916 ISSN 0936-5214 © Thieme Stuttgart · New York

916
A. Egger, C. J. Leumann
LETTER
(7) All compounds 2-9 were fully characterized by 1H, 13C (31P)
NMR, IR, UV, and MS. Analytical data of 8: 1H NMR (300
MHz, CDCl3): 1.01, 1.03, 1.06, 1.08, 1.10 (5s, 12H,); 1.42 (d,
J = 16.9, 1H); 1.75 - 1.86 (m, 1H); 1.88 (s, 3H); 2.07, 2.15
(2dd, J = 12.9, 2.6, 1H); 2.52 - 2.61 (m, 2H); 2.70 - 2.80 (m,
1H); 3.56 - 3.48 (m, 2H); 3.63 - 3.71 (m, 2H); 3.79 (s, 6H);
3.82 - 3.89 (m, 1H); 3.95 (d, J = 5.2, 1H); 3.99 (s, 1H); 4.07-
4.18 (m, 2H); 4.35, 4.41 (2m, 1H); 4.98, 5.09 (2 s, 2H); 6.85 -
6.90 (m, 4H); 6.95, 7.04 (2d, J < 0.1, 1H); 7.23 - 7.48 (m, 9H);
8.29 (bs, 1H). 31P NMR (161.9 MHz, CDCl3): 154.02, 154.55.
LSIMS-MS: 828 (M+H+)+. Analytical data of 9: 1H NMR (300
MHz, CDCl3): 0.95, 0.97, 1.00, 1.02, 1.02, 1.03, 1.05, 1.05
(8s, 12H); 1.41, 1.52 (2d, J = 14.5, 1H); 1.74 - 1.85 (m, 1H);
2.06, 2.17 (2dd, J = 13.2, 3.6, 1H); 2.49 - 2.54 (m, 2H); 2.69 -
2.72 (m, 1H); 3.29 - 3.41 (m, 2H); 3.53 - 3.70 (m, 2H); 3.80 (s,
6H); 3.81 - 3.86 (m, 1H); 3.95, 4.01 (2d, J = 5.2, 1H); 4.15 -
4.22 (m, 2H); 4.35, 4.41 (2d, J = 9.9, 1H); 4.41, 4.55, 4.56,
4.64 (4d, J = 15.8, 2H); 5.12, 5.42 (2s, 1H); 6.88 - 6.94 (m,
4H); 7.25 - 7.61 (m, 12H); 7.99 - 8.01 (m, 2H); 8.11, 8.16 (2s,
1H); 8.71, 8.73 (2s, 1H); 9.05, 9.06 (2bs, 1H). 31P NMR (161.9
MHz, CDCl3): 154.28, 154.40. LSIMS-MS: 942 [M+H+]+.
(8) Saenger, W. Principles of Nucleic Acid Structure; Springer,
1984.
(9) Marky, L. A.; Breslauer, K. J. Biopolymers, 1987, 26, 1601.
(10) For a comparison of CD-spectra of DNA and enantio-DNA in
the case of the sequence d(CGCGCG) see: Urata, H.,
Shinohara, K., Ogura, E., Ueda, Y., Akagi, M. J. Am. Chem.
Soc. 1991, 113, 8174.
(11) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science
1991, 254, 1497.
(12) Diederichsen, U. Angew. Chem., Int. Ed. Engl. 1996, 35, 445.
Diederichsen, U. Angew. Chem., Int. Ed. Engl. 1997, 36,
1886.
Article Identifier:
1437-2096,E;1999,0,S1,0913,0916,ftx,en;W07399ST.pdf
Synlett 1999, S1, 913–916 ISSN 0936-5214 © Thieme Stuttgart · New York