yDNA: A new geometry for size-expanded base pairs

Article abstract of DOI:10.1002/anie.200461036

Room for expansion: Insertion of a benzene ring into the natural adenine heterocycle to generate the analogue yA results in the formation of base pairs that are 2.4 A wider than natural ones (see structure). Melting temperature and free energy data show t

Full text of DOI:10.1002/anie.200461036

Communications
DNA Structures
yDNA: A New Geometry for Size-Expanded
Base Pairs**
Haige Lu, Kaizhang He, and Eric T. Kool*
Previous studies have explored whether it is possible to
replace the information-encoding part of DNA—the bases
and base pairs—with other molecular replacements, and
whether such new base pairs might function in recognition
and in replication processes, two of the defining character-
istics of the natural genetic system.[1–6] Many of these studies
have been aimed at designing base pairs that can function
within the content of the natural genetic system, for
expansion of natureꢀs genetic alphabet. However, recent
studies have moved beyond the purine–pyrimidine frame-
work of the natural system. Examples of this new approach
include metal-bridged base pairs,[4] nonpolar base pairs,[2,4]
and pairs containing more than the three different types of
hydrogen bonds found in nature.[5] We recently adopted a
different strategy and described a molecular design in which
the dimensions of the natural pairs are stretched by insertion
of a benzene ring into the natural heterocycles, thus rendering
base pairs 2.4 Š wider than the natural ones.[6] Such broad
alterations of the natural design are not necessarily expected
to be compatible with the natural genetic system. Rather, such
studies are useful as a test of researchersꢀ understanding of
how genetic systems function in general, and they may also
lead to useful new applications in biotechnology and in self-
assembling nanostructures.
Herein we evaluate whether size-expansion of DNA base
pairs can be carried out using a new geometry. Our previous
design of expanded DNA-like bases (xDNA) involved a
linear extension of pyrimidines and purines by addition of a
benzene ring to each of the four DNA base heterocycles.[6]
However, examination of models suggested that at least one
other design strategy involving benzohomologation also
appeared to allow for reasonable base-pair geometries and
stacking with neighboring pairs. This new design, termed
“yDNA” (an abbreviated form of “wide DNA”; Scheme 1)
involves a different extension vector, but yields a similar
degree of perturbation from the framework formed by the
natural pair. Analogous designs for the other three nucleo-
bases can be envisioned (not shown). In this initial study we
Scheme 1. a) Structures of the free deoxyribonucleosides yA and natu-
ral A; b) proposed base-pair pattern for yA·T (left), compared with the
natural A·T base pair (right). The arrow shows the extension vector for
the yA base which adds about 2.4 Š to the length of natural adenine
base.
tested the concept with one example, that of the expanded
adenine base analogue (yA) and the corresponding deoxy-
riboside (dyA), both of which were previously unknown.
We developed a synthetic route to prepare the phosphor-
amidite derivative of the new nucleoside analogue dyA in
10 steps (6% overall yield on a gram scale, Scheme 2).
Methylation of indole 1 at the 5 position was achieved by
addition of three equivalents of methylmagnesium chloride to
the unprotected indole 1, followed by ring oxidation to
restore aromaticity.[7] The transformation of methylindole 3[8]
to indole-5-carboxaldehyde 4 was a key step; tris(dimethyla-
mino)methane converted 3 into the enamine intermediate
which was then oxidized by KMnO4 in one pot to afford
compound 4.[9] The protected 2’-deoxyriboside 7 was formed
in three subsequent steps from the indole-5-aldehyde 4.
Conventional methods were then used to convert the
deoxyriboside 7 into the amidine-protected dyA phosphor-
amidite 11.
Studies showed that the yA base is fluorescent, where as
its natural congener is not. The yA free nucleoside was found
to have absorption maxima at 262 and 355 nm in methanol. A
fluorescence spectrum in methanol revealed that the nucleo-
side emits blue fluorescence (lmax = 433 nm) with a quantum
yield of 0.12 (Figure 1). Thus the yA fluorescence is red-
shifted by about 40 nm relative to the xA analogue.[6]
We prepared a short oligomer of sequence 5’-TyAT-3’ to
confirm that dyA could be incorporated intact into a DNA by
automated DNA synthesis. The expected trimer structure was
confirmed by 1H NMR spectroscopy and electron spray
ionization (ESI) mass spectrometry after deprotection with
concentrated ammonium hydroxide and partial purification
by dialysis.
[*] H. Lu, Dr. K. He, Prof. Dr. E. T. Kool
Department of Chemistry
Stanford University
Stanford, CA 94305-5080 (USA)
Fax: (+1)650-725-0259
E-mail: kool@leland.stanford.edu
[**] This work was supported by the U.S. National Institutes of Health
(GM63587) and by an Abbott Laboratories Graduate Research
Fellowship to H.L.
We then investigated the stacking and pairing properties
of dyA in the context of natural DNA oligonucleotides.
Thermodynamic data from thermal denaturation studies
(Table 1) showed that yA (like xA)[6] stacks more strongly
than natural A at the terminus of the duplex formed by 5’-
Supporting information for this article (synthesis and structure
characterization, thermal denaturation curves, CD spectra) is
available on the WWW under http://www.angewandte.org or from
the author.
5834
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200461036
Angew. Chem. Int. Ed. 2004, 43, 5834 –5836

Angewandte
Chemie
Scheme 2. Reagents and conditions: a) 1. CH3MgCl,THF,À258C; 2. Pb(OAc)4, CH2Cl2, 42%; b) 1. NaOH, CH2Cl2; 2. ClSO2Ph, 87%; c) 1. DMF,
CH(NMe2)3, 1058C; 2. KMnO4, THF/H2O 10:1, 70%; d) Na2S2O4, THF/H2O=2:1; e) MeCONMe2, guanidine carbonate, 1508C, 61% over two
steps; f) 1. NaH, CH3CN, RT; 2. Hoffer’s chlorosugar, 08C, 92%, b/a>20:1; g) NaOMe, MeOH, 91%; h) CH(OMe)2NMe2, pyridine; i) DMTrCl,
pyridine, DIPEA, 62% over two steps; j) chlorodiisopropyl-cyanoethylphosphoramidite, CH2Cl2, DIPEA, 70%. DMTr=4,4’-dimethoxytriphenyl-
methyl, DIPEA=N,N-diisopropylethylamine, Tol=tolyl.
Although the yA·T pair was found to be destabilizing in
natural DNA, probably because of backbone distortions, we
expected that such strain might be avoided if all pairs were
expanded in a homologous way. Thus, further studies were
performed to test whether it is possible to construct fully
expanded helices composed of yA·Tand T·yA base pairs. This
was tested in two contexts. The first one contained a self-
complementary sequence (Table 1, entry 16) and the second
one (Table 1, entry 18) contained a non-self-complementary
pair of oligonucleotides. Both assays showed distinct cooper-
ative transition curves in thermal denaturation experiments
(see Supporting Information). In the first case, the non-
natural helix displayed a higher melting temperature (Tm, by
258C) and more favorable free energy (by 4.1kcalmol À1 at
378C) relative to the natural helix of the analogous sequence.
Similarly, the helix formed with the second sequence gave a
Tm value 238C higher and a free energy 4.9 kcalmolÀ1 higher
than the natural helix. We hypothesize that the added stability
of these helical complexes arises from the strong stacking of
Figure 1. Emission and excitation spectra of the free nucleoside yA in
the larger base pairs; further experiments will be needed to
better understand this.
methanol.
Circular dichroism spectra of the duplex formed in the
second case (Table 1, entry 18) suggest an overall helical form
resembling that of B-DNA of the analogous sequence (see
Supporting Information), but with positive and negative
bands red-shifted by 5–20 nm, presumably as a consequence
of the extended p system of yA. Interestingly, the single
component strands of the putative yDNA duplex also show
similar spectra, which is indicative of strong helical stacking in
the single-stranded state.
Our study has demonstrated that the new geometric
design for benzo-fused bases is successful for at least one such
base analogue (yA), and the data show that yA·T and T·yA
pairs can self-assemble into cooperative helices that are
considerably more stable than natural DNA. Thus, yDNA
CGCGCG-3’. This result may be a consequence of the
expanded surface area and higher hydrophobicity of the new
analogue. We then tested the pairing properties of yA
opposite natural bases in DNA (Table 1, entries 4–15). With
a single substitution in DNA, yAwas found to be destabilizing
to some extent, which may be a result of the distortion of the
DNA backbone caused by the increased size of yA relative to
A. Interestingly, when yA was paired with T, the duplex was
less destabilized by 1.5–2.1 kcalmolÀ1 than when yA was
paired opposite A,G, and C. This result suggests that, despite
some destabilization, yA may be able to form a selective
hydrogen-bonded pair with T in the natural duplex.
Angew. Chem. Int. Ed. 2004, 43, 5834 –5836
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5835

Communications
Table 1: Melting temperatures and free energies for DNA helices containing yA bases and base pairs.
[1] J. A. Piccirilli, T. Krauch, S. E.
Moroney, S. A. Benner, Nature
1990, 343, 33 – 37.
Duplex[a]
Tm [8C][b]
ÀDG3o 7 [kcalmolÀ1 [c]
]
1
2
3
4
64.4
52.6
43.8
38.7
13.2Æ0.5
10.2Æ0.1
9.2Æ0.4
8.8Æ0.1
[2] a) B. A. Schweitzer, E. T. Kool, J.
Am. Chem. Soc. 1995, 117, 1863 –
1872; b) T. J. Matray, E. T. Kool, J.
Am. chem. Soc. 1998, 120, 6191 –
6192; c) T. J. Matray, E. T. Kool,
Nature 1999, 399, 704 – 708;
d) D. L. McMinn, A. K. Ogawa, Y.
Wu, J. Liu, P. G. Schultz, F. E.
Romesberg, J. Am. Chem. Soc.
1999, 121, 11585 – 11586; e) E. L.
Tae, Y. Wu, G. Xia, P. G. Schultz,
F. E. Romesberg, J. Am. Chem.
Soc. 2001, 123, 7439 – 7440; f) T.
Mitsui, A. Kitamura, M. Kimoto, T.
To, A. Sato, I. Hirao, S. Yokoyama,
J. Am. Chem. Soc. 2003, 125, 5298 –
5307.
5
29.2
31.0
32.2
33.9
35.2
25.9
27.5
28.0
27.5
41.4
35.4
44.7
19.5
44.1
20.9
6.8Æ0.1
7.1Æ0.1
7.3Æ0.1
7.7Æ0.1
7.9Æ0.1
5.8Æ0.1
6.3Æ0.1
6.4Æ0.1
6.3Æ0.1
9.5Æ0.1
8.0Æ0.1
8.7Æ0.1
4.6Æ0.2
9.5Æ0.3
4.6Æ0.2
6
7
8
9
10
11
12
13
14
15
16
17
18
19
[3] H. P. Rappaport, Nucleic Acids
Res. 1988, 16, 7253 – 7267.
[4] a) E. Meggers, P. L. Holland, W. B.
Tolman, F. E. Romesberg, P. G.
Schultz, J. Am. Chem. Soc. 2000,
122, 10714 – 10715; b) K. Tanaka,
A. Tengeiji, T. Kato, N. Toyama, M.
Shionoya, Science 2003, 299, 1 21 2 –
1213.
[5] N. Minakawa, N. Kojima, S. Hikish-
ima, T. Sasaki, A. Kiyosue, N.
Atsumi, Y. Ueno, A. Matsuda, J.
Am. Chem. Soc. 2003, 125, 9970 –
9982.
[6] H. Liu, J. Gao, S. R. Lynch, Y. D.
Saito, L. Maynard, E. T. Kool, Sci-
ence 2003, 302, 868 – 871.
[7] J. Quick, B. Saha, Tetrahedron Lett.
1994, 35, 8553 – 8556.
[8] O. Ottoni, R. Cruz, R. Alves, Tet-
rahedron 1998, 54, 13915 – 13928.
[9] a) R. D. Clark, D. B. Repke, Het-
erocycles 1984, 22, 195 – 221;
b) E. C. Riesgo, X. Jin, R. P. Thum-
mel, J. Org. Chem. 1996, 61, 3017 –
3022.
[a] f in entries 8 and 13 denotes an abasic tetrahydrofuran site. [b] Entries 1, 2, 15, 16: the buffer
contained 1m NaCl and 10 mm Na2HPO4 at pH 7.0; DNA concentration: 5.0 mm; error in the Tm value:
Æ0.58C or less. Entries 3–14, 17, 18: buffer contained 100 mm NaCl, 10 mm MgCl2, and 10 mm sodium
1,4-piperazinediethanesulfonate at pH 7.0; DNA concentration: 5.0 mm. Error in the Tm value: Æ0.58C
or less. [c] Data were obtained by averaging free energies from curve fits and a van’t Hoff plot. The
van’t Hoff data were obtained by plotting 1/Tm versus ln(C/4) (where C is the total oligonucleotide
concentration) with data from five concentrations. Curve fits data were averaged from fits of five
denaturation curves.
[10] H. Liu, J. Gao, L. Maynard, Y. D.
Saito, E. T. Kool, J. Am. Chem. Soc.
2004, 126, 1102 – 1109.
may be a candidate for a new genetic system, distinct from the
recently described xDNA design.[6,10] Like xDNA, it appears
that the sugar–phosphate backbone in yDNA can tolerate the
stretched size of bases and the altered pairing direction, while
retaining the structure of the duplex. However, beyond the
hypothesized base pairing, the structure of yDNA is unknown
and needs to be studied. Also unknown is whether other
analogously expanded bases and base pairs could be con-
structed; studies are underway to test these.
Received: June 22, 2004
Keywords: biomimetic synthesis · DNA structures ·
.
hydrogen bonds · nucleosides
5836
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 5834 –5836