Base-base interactions in the minor groove of double-stranded DNA

Article abstract of DOI:10.1021/jo052194c

A method has been developed for the synthesis of bisheaded nucleosides with thymine and adenine base moieties. We have demonstrated that, when incorporated in oligonucleotides, extrahelical A-T base interactions are possible when the bisheaded nucleosides

Full text of DOI:10.1021/jo052194c

VOLUME 71, NUMBER 15
JULY 21, 2006
© Copyright 2006 by the American Chemical Society
Base-Base Interactions in the Minor Groove of Double-Stranded
DNA
Tongfei Wu, Koen Nauwelaerts, Arthur Van Aerschot, Matheus Froeyen,
Eveline Lescrinier, and Piet Herdewijn*
Laboratory of Medicinal Chemistry, Rega Institute for Medical Research, Minderbroedersstraat 10,
B-3000 LeuVen, Belgium
Piet.Herdewijn@rega.kuleuVen.ac.be
ReceiVed October 21, 2005
A method has been developed for the synthesis of bisheaded nucleosides with thymine and adenine base
moieties. We have demonstrated that, when incorporated in oligonucleotides, extrahelical A-T base
interactions are possible when the bisheaded nucleosides are positioned in opposite strands of the duplex
and are separated from each other by one regular base pair.
Introduction
recognition of double-stranded DNA with heterocyclic polya-
mides1 or with triple helix forming oligonucleotides.2 Presenting
bases outside the helix has its precedent in biochemistry (base
flipping for nucleic acid methylation),3 in structural biology
Two strands of complementary DNA form a helical duplex
in which the base pairs form the core of the duplex, and these
pairs are accessible from the outside via the major and minor
grooves. These grooves can be used for sequence-specific
(1) Dervan, P. B.; Edelson, B. S. Curr. Opin. Struct. Biol. 2003, 13,
284-299.
(2) Praseuth, D.; Guieyesse, A. L.; He´le`ne, C. Biochim. Biophys. Acta.
1999, 10, 181-206.
* To whom correspondence should be addressed. Phone: +32-16-337387.
Fax: +32-16-337340.
10.1021/jo052194c CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/20/2006
J. Org. Chem. 2006, 71, 5423-5431
5423

Wu et al.
FIGURE 1. Nucleosides with two base moieties of the TA and TT
motif.
(such as the formation of bulge loops),4 as a spontaneous event
(the flipped-out base can be trapped using â-cyclodextrin),5 and
in the field of synthetic nucleic acid chemistry.6 In the latter
example, two base moieties on the same backbone motif are
involved in the formation of a stable duplex (i.e., one base is
involved in base pairing within the double helix, while the
second base is sticking out in the minor groove).
The structure repertoire of nucleic acids could be expanded,
and the specificity of recognition between two oligonucleotides
could be increased, when base pairing outside the double helix
could be realized. Extending the genetic information system in
regular DNA is usually done by increasing the number of
specific base pairs7 and not by increasing the number of
canonical base pairs involved in the recognition process.
Additional base pair recognition might be realized by introduc-
tion of bisheaded nucleotides in complementary DNA strands
which can interact with each other in the grooves of the duplex.
To evaluate the potential of this recognition system, we
synthesized an oligonucleotide based on a TA and TT motif (XY;
X-base involved in base pairing within the double helix;
Y-base presented at the outside of the helix) (Figure 1). The
second base is connected at a 4′-C-methyl group of the
nucleoside. A model shows that base pairing within the minor
groove might be possible between a 4′-C-purine and a 4′-C-
pyrimidine base, provided that the modified nucleotides are
introduced within opposite strands of the double-stranded DNA
and separated from each other by one regular base pair (Figure
2).
FIGURE 2. Potential for base pairing within the minor groove of
double-stranded DNA using complementary bisheaded nucleosides.
The synthetic route to 4′-C-[(N6-benzoyladenin-9-yl)methyl]-
3′-O-[2-cyanoethoxy-(diisopropyl) aminophosphinyl]-5′-O-
monomethoxytritylthymidine is outlined in Scheme 1.
The 3′-O-(tert-butyldimethylsilyl)-4′-(hydroxymethyl)-thy-
midine 1 was synthesized according to a previously described
procedure.10 Both primary hydroxyl groups of 1 were converted
into their triflate by using trifluoromethanesulfonyl chloride in
pyridine, giving compound 2 in 85% yield. The adenine base
was introduced at the 4′-position using NaH and adenine in dry
DMF at 50 °C in 64% yield. 2D NMR (gHMBC) was used to
confirm the right connection between sugar and nucleobase
moieties. The coupling between the 5′-H of sugar and the 2-C
of thymine proved the formation of a 5′,O2-anhydrobond. The
protons of the 4′-methylene group were coupled with both the
C-4 and C-8 of the adenine base.
Hydrolysis of the anhydride using 2 N NaOH in dioxane
yielded compound 4 in 46% yield. The 5′-OH of compound 4
was protected with a TBDMS group using TBDMSCl and
imidazole in acetonitrile in 58% yield. Protection of the
exocyclic amino group of the adenine base was achieved using
an excess of benzoyl chloride in pyridine and subsequent
treatment with sat. NH3/MeOH at 0 °C for half an hour. The
desired monobenzoyl-protected amine 6 was obtained in 94%
yield. Both TBDMS groups were removed using TBAF, giving
compound 7 in 57% yield. Reaction of compound 7 with
MMTrCl in anhydrous pyridine afforded 8 in 81% yield. Finally,
the secondary hydroxyl group of 8 was phosphitylated to obtain
the desired phosphoramidite 9, which was then used as a
building block for oligonucleotide synthesis.
Results and Discussion
Synthesis. Synthetic chemistry for the introduction of a
functionalized alkyl group in the 4′-position of a nucleoside has
been described previously,8,9 and our slightly modified procedure
is based on this chemistry.
(3) Allan, B. W.; Beechem, J. M.; Lindstrom, W. M.; Reich, N. O. J.
Biol. Chem. 1998, 273, 2368-2373.
(4) Shen, L. X.; Cai, Z.; Tinoco, I., Jr. FASEB J. 1995, 9, 1023-1033.
(5) Spies, M. A.; Schowen, R. L. J. Am. Chem. Soc. 2002, 124, 14049-
14053.
(6) Wu, T.; Froeyen, M.; Schepers, G.; Mullens, K.; Rozenski, J.; Busson,
R.; Van Aerschot, A.; Herdewijn, P. Org. Lett. 2004, 6, 51-54.
(7) Geyer, C. R.; Battersby, T. R.; Benner, S. A. Structure 2003, 11,
1485-1498.
(8) Maag, H.; Schmidt, B.; Rose, S. J. Tetrahedron Lett. 1994, 35, 6449-
6452.
Following a similar synthetic strategy for obtaining compound
9, 4′-C-(thymin-1-yl)methyl-5′-monomethoxytrityl-3′-O-[2-cya-
(9) Fensholdt, J.; Thrane, H.; Wengel, J. Tetrahedron Lett. 1995, 36,
2535-2538.
(10) O-Yang, C.; Wu, H. Y.; Fraser-Smith, E. B.; Walker, K. A. M.
Tetrahedron Lett. 1992, 33, 37-40.
5424 J. Org. Chem., Vol. 71, No. 15, 2006

Bisheaded DNA
SCHEME 1. Synthesis of the TA Building Block
noethoxy diisopropylaminophosphinyl]-thymidine was obtained
as outlined in Scheme 2.
MMTrCl in anhydrous pyridine afforded 13 in 83% yield.
Finally, phosphitylation of compound 13 gave the desired
phosphoramidite 14.
Substitution of 2 using NaH and thymine in dry DMF at 50
°C gave compound 10, which was treated with 2 N NaOH in
dioxane. This compound was difficult to purify using column
chromatography due to comigration of an impurity. Therefore,
preparative TLC was used to obtain a pure sample. The 2D
NMR spectrum (gHMBC) showed that the protons on the extra
methylene group at 4′-C of 11 are coupled with C-2 and C-6 of
the newly introduced thymine base of compound 11, proving
the desired connection between thymine and methylene group
on 4′-C. The right connection between N1 of the newly
introduced thymine and carbon 4′-C is characterized by the peak
at 46.92 ppm in the 13C NMR spectrum. NMR (1H and 13C)
can be used to distinguish between a C-N and C-O bond in
the alkylation reaction of pyrimidine bases (for example, see
ref 11).
With the phosphoramidites 9 and 14 at hand, several
oligonucleotides were synthesized comprising single or triple
incorporations of the modified thymidine nucleoside analogues.
The assembly was performed using the phosphoramidite ap-
proach. The nucleoside analogues were used at 0.07 M, and
coupling was allowed to proceed for 12 min. Ethylthiotetrazole
(ETT) was used as activator, resulting mostly in good coupling
yields (>95%). The synthesized oligomers were isolated fol-
lowing ion exchange chromatography at pH 12 and desalting
by gel filtration. The alkaline pH ensures denaturing conditions
and allows straightforward isolation of the desired peak in >90%
purity. All synthesized oligonucleotides were analyzed by mass
spectrometry (LC/MS, see the Experimental Section) to prove
the correct incorporation of the nucleotide analogues, and the
obtained data are shown in Table 1. After submission of this
manuscript, an article was published describing the incorporation
of a 2′-C-(2-(thymine-1-yl)ethyl nucleoside in an LNA-modified
oligonucleotide and studying its influence on the stability of
three-way junctions.12
The TBDMS group of compound 11 was removed using
tetrabutylammonium fluoride. Compound 12 was obtained in
12% overall yield from compound 2. Tritylation of 12 with
(11) Hossain, N.; Rozenski, J.; De Clercq, E.; Herdewijn, P. J. Org.
Chem. 1997, 62, 2442-2447.
J. Org. Chem, Vol. 71, No. 15, 2006 5425

Wu et al.
SCHEME 2. Synthesis of the TT Building Block
TABLE 1. ESI-MS Monoisotopic Data of Oligonucleotides
Containing Modified Monomeric Units
TABLE 2. Influence of the Incorporation of Several Modifications
(TT and TA) on the Stability of DNA Duplexesa
DNA sequences
mass calcda
ESI-MS
entry
1
sequences
Tm
∆Tm
∆Tm/mod
d(3′-CGC TATATATACGC-5′)
d(3′-GCGTTAAGCG-5′)
d(3′-CGCTATACGC-5′)
d(3′-CGCTTTACGC-5′)
d(3′-GCGTAAAGCG-5′)
3556.7
2900.6
2820.5
2811.5
2909.6
3556.7
2900.6
2820.5
2811.5
2909.5
d(3′-CGC TTTTCGC-5′)
d(5′-GCG AAAAGCG-3′)
d(3′-CGC TATATATACGC-5′)
d(5′-GCG A A A A GCG-3′)
d(3′-GCGTAAGCG-5′)
d(5′-CGCATT CGC-3′)
d(3′-GCGTAAAGCG-5′)
d(5′-CGCA TT CGC-3′)
d(3′-GCGTTAAGCG-5′)
d(5′-CGCA TT CGC-3′)
d(3′-GCGTAA GCG-5′)
d(5′-CGCATTA CGC-3′)
d(3′-GCGTAA GCG-5′)
d(5′-CGCATTT CGC-3′)
d(3′-GCGTTAAGCG-5′)
d(5′-CGCATTA CGC-3′)
d(3′-GCGTTAAGCG-5′)
d(5′-CGCATTT CGC-3′)
d(3′-GCGTAAAGCG-5′)
d(5′-CGCATTA CGC-3′)
46.4
2
45.2
40.7
30.0
39.0
34.7
38.5
35.3
36.3
37.5
-1.5
-0.4
3
a Calculated value on average mass.
4
-10.7
-1.7
-6.0
-2.2
-5.4
-4.4
-3.2
-10.7
-1.7
-6.0
-2.2
-2.7
-2.2
-1.6
5
Thermal Stability Study. The phosphoramidites 9 and 14
were used to incorporate the 4′-branched thymidines TA and
TT into several oligonucleotides. In the first sequence (3′-
CGCTATATATACGC-5′), we incorporated four consecutive TA
moieties. This allowed us to evaluate the influence of the sugar
moiety versus an acyclic backbone on dsDNA stability. When
using four consecutive acyclic monomers as described before,6
the ∆Tm/mod was as high as -4.0 °C. However, when using
four TA analogues, the ∆Tm/mod amounted to only -0.4 °C
(Table 2 entries 1 and 2). A considerable increase in duplex
stability therefore is observed when replacing the acyclic
backbone by a 4′-C-branched furanose backbone. Recently,
Wengel et al. reported that introduction of 4′-C-(N-methylpip-
erazino)methyl-substituted nucleosides into DNA significantly
increased the thermal stability of a DNA duplex,13 due to the
possibility of salt formation of the piperazine moiety. In contrast
to this 4′-C-(N-methylpiperazino)methyl moiety, the base (A)
connected at the 4′-C-methyl group is not expected to be
protonated at physiological pH. With a single incorporation,
however, there was a detrimental loss in affinity, albeit still
smaller than with the acyclic monomers (entries 4 and 6). With
the thymine base at the branching point (TT), the destabilization
6
7
8
9
10
a Tm was determined at UV 260 nm in NaCl (0.1 M), KH2PO4 (20 mM,
pH 7.5), EDTA (0.1 mM). Entries are in °C.
effect remained small (entries 5 and 7). This difference between
a purine and pyrimidine base could not been explained.
To try to prove the hypothesis that base pair interaction might
occur within the minor groove of dsDNA, TA and TT were
incorporated into both strands of the dsDNA sequence d(3′-
GCGTAAGCG-5′)(5′-CGCATTCGC-3′). In each case, as re-
quired by the model, the modified nucleotides were separated
by a regular A-T base pair (Table 2, entries 8-10). The drop
in Tm varied between -3.2 and -5.4 °C (or -1.6 and -2.7
°C/mod) and thus is lower than the destabilization caused by a
single TA moiety (but higher than a single TT moiety), indicating
a stabilizing effect by introduction of the TT nucleotide in the
opposite strand of the TA nucleotide. It is rewarding to notice
that the Tm rises 0.6 °C when a second modification is introduced
(12) Pedersen, S. L.; Nielsen, P. Org. Biomol. Chem. 2005, 3, 3570-
3575.
(13) Raunkjaer, M.; Bryld, T.; Wengel, J. Chem. Commun. 2003, 13,
1604-1605.
5426 J. Org. Chem., Vol. 71, No. 15, 2006

Bisheaded DNA
FIGURE 3. Numbering of duplex 8.
FIGURE 5. Part of the NOESY spectrum (mixing time 150 ms) of
A
duplex 8. Indicated are some interesting cross-peaks between H2(T6
)
T
and proton signals of the opposite strand. (A) H1′(T6 ), (B) H1′(A14),
T
(C,D) H6′/H6′′(T15T), (E) CH3(T15 extra base moiety).
FIGURE 4. Overlay of the aromatic gHSQC spectra of single-stranded
d(5′-CGCATTA CGC-3′) (drawn in green) and the aromatic gHSQC
spectrum of construct 8 (drawn in red). The boxed resonance signals
originate from the H2 proton of the extra adenine base moiety.
in the opposing strand, hinting to a T-A interaction in the minor
groove (entry 6 versus 8). However, a minor groove A-A
interaction looks even much stronger in the studied sequence
context, as the Tm increases from 30.0 and 34.7 °C, respectively,
to 37.5 °C when a TA block is incorporated in both opposing
strands (entries 4 and 6 versus 10). The analogous TT incorpora-
tion has no positive effect on duplex stabilities, as here the Tm
is lowered further compared to a single TT incorporation (entries
5 and 7 versus 9). It is interesting to observe that introduction
of identical bisheaded bases (entries 9 and 10) gives less
destabilization than introduction of different bisheaded bases
(entry 8) in the opposite strands, and hydration of the minor
groove might play a role in this phenomenon.
It can be concluded that the studied interaction of TT with
TA or TA with TA in opposite strands separated by one regular
base pair leads to a duplex stabilization when compared to
incorporation of a single TA unit. The reason for the stabilization
effect is not clear. The 4′-C-methylthymine or 4′-C-methylad-
enine substituent can be situated outside the helix or within the
minor groove. Interactions in the grooves are possible with the
information present in the floor of the groove or by base pairing
or stacking interactions of the extra nucleobases.
Therefore, we started an NMR investigation on two repre-
sentative examples (i.e., the T-A interaction of the duplex of
entry 8 and the position of the T base in the duplex of entry 5).
Duplex 8 was selected (instead of the more stable duplex 10)
for structural studies because of the potential for Watson-Crick
base pairing within the minor groove.
NMR Analysis and Molecular Modeling of Duplexes 5 and
8: Localization of the Extrahelical Bases. To shed light on
the structural mechanisms that stabilize the above duplexes in
which TA and TT building blocks are present in opposite strands,
a preliminary NMR study was undertaken. As a model to study
these interactions, duplex 8 (Figure 3) was selected because the
extra bases (A and T) could interact as well by base pairing as
by stacking. As discussed above, it is worth noting that the
introduction of a TT building block in construct 6 increases the
stability of the resulting duplex (construct 8) by 0.6 °C.
First NOESY, TOCSY, and H1-C13 gHSQC spectra were
recorded for both single-stranded molecules. Following titration
FIGURE 6. Averaged (A) and detailed (B) structure of the molecular
modeling experiment showing the interaction of the T and A bases in
the minor groove of the duplex. Figure B also shows the observed NOE
contacts in the modified region.
of both strands to obtain an equimolar solution, spectra of the
double-stranded molecule were recorded. It was encouraging
to notice that the resonance signals for the additional bases did
shift significantly following the formation of the duplex
structure, indicating structural changes within the modified
region. The proton and carbon signal of the H2 proton in the
extra adenine base moiety was shifted from 8.04 ppm (H1) and
155.12 ppm (C13) in the single strand to 7.77 ppm (H1) and
154.85 ppm (C13) in the duplex (Figure 4). The methylene signal
of the extra thymine base moiety also shifted from 1.82 ppm
(H1) to 1.73 ppm (H1).
Considering these promising initial results, partial assignment
of the resonance signals in the modified region was started using
the gHSQC, TOCSY, and NOESY spectra of the duplex. Some
interesting NOE contacts could be observed between the adenine
proton H2(T6A) and proton signals of the opposite strand [H1′-
(A14), H6′/H6′′(T15T), and CH3(T15T extra base moiety)] as
is indicated in Figures 5 and 6B.
This pattern of NOE cross-peaks is not corresponding to the
expected pattern in a canonical AT base pair where no cross-
J. Org. Chem, Vol. 71, No. 15, 2006 5427

Wu et al.
column chromatography on a silica gel column (n-hexane/EtOAc,
4:1, 2:1) to afford compound 2 (3.08 g, 4.73 mmol, 85%) as a
yellow amorphous solid. H NMR (500 MHz, CDCl3) δH 0.12 (s,
peak is expected between the methyl protons of the thymine
base and the H2 proton of adenine. The observed pattern better
fits a situation in which the two bases are stacked upon each
other in the minor groove. In this case, an NOE contact cross-
peak is expected as well between H2 and the H6′/H6′′ protons
(cross-peaks C and D, Figure 5) as between H2 and the methyl
group (cross-peak E, Figure 5). This positioning of the bases
also explains the NOE contacts with the H1′ protons of residues
A13 and A14.
On the basis of the available NMR restraints, a model was
built using the Amber software and the energy was minimized
in Sander. The duplex was solvated in a water box, and
counterions were added followed by molecular dynamics
simulation with all restraints removed (500 ps at 300 K). The
results of the modeling experiments are in agreement with the
observed NOE effects. This model demonstrates that the
extrahelical bases (A and T) are situated in the minor groove.
The interaction, however, is not specific (no Watson-Crick TA
base pairing), and the stabilization of the interaction is mainly
of hydrophobic nature.
To evaluate if the extrahelical T base is occupying a minor
groove position when no A base is present in the opposite strand
(preorganization for base-base interactions), we analyzed the
structure of duplex 5 (Table 2) by NMR. The signals in the
NMR spectra of the extra thymine base, however, do not really
shift when the spectra of the single-strand TT-oligo is compared
with the spectra of the double-stranded oligo. No NOE effects
are observed between the extrahelical thymine base and the
protons of the regular DNA duplex (except for NOE contact
between protons of the extrahelical thymine base and its own
sugar moiety). These signals are broader than these of the NOE
signals of bases involved in base pairing in the center of the
helix. Therefore, we may conclude that the extrahelical thymine
base has a high degree of conformational freedom when no
adenine base is present in the opposite strand.
1
3H, SiCH3), 0.13 (s, 3H, SiCH3), 0.91 (s, 9H, CH3), 1.93 (s, 3H,
T-CH3), 2.40-2.50 (m, 1H, 2′-Ha), 2.68-2.73 (m, 1H, 2′-Hb), 4.55
(d, J ) 11.1 Hz, 1H, 5′-Ha), 4.59 (d, J ) 10.6 Hz, 1H, 6′-Ha), 4.68
(d, J ) 10.6 Hz, 1H, 6′-Hb), 4.74 (d, J ) 11.1 Hz, 1H, 5′-Hb), 4.83
(t, J ) 6.2 Hz, 1H, 3′-H), 5.91 (t, J ) 6.2 Hz, 1H, 1′-H), 7.04 (s,
1H, 6-H), 8.81 (br s, 1H, NH); 13C NMR (500 MHz, CDCl3) δC
-5.3, -4.7 (Si(CH3)2), 12.2 (T-CH3), 17.8 (C(CH3)3), 25.5
(C(CH3)3), 40.1 (C-2′), 73.6 (C-5′ or C-6′), 73.7 (C-3′), 73.9 (C-6′
or C-5′), 84.5 (C-4′), 89.2 (C-1′), 111.6 (C-5), 118.5 (q, J ) 318
Hz, CF3), 126.7 (q, J ) 366 Hz, CF3), 137.7 (C-6), 149.7 (C-2),
163.4 (C-4); ESI HRMS calcd for C19H29F6N2O10S2Si [M + H]+
651.0937; found 651.0922.
4′-C-[(Adenin-9-yl)methyl]-3′-O-[(1,1-dimethylethyl)dimeth-
ylsilyl]-O2,5′-anhydrothymidine (3). To a solution of adenine (0.61
g, 4.62 mmol) in 45 mL of dried DMF was added NaH (80%, 345
mg, 11.6 mmol) at room temperature. The mixture was heated to
70 °C and kept stirring for 10 min. The reaction mixture was cooled
to 50 °C, and a solution of compound 2 (3.0 g, 4.62 mmol) and a
catalytic amount of 18-crown-6 in 10 mL of DMF were added.
After being stirred for 4 h, the solvent was removed in vacuo. The
residue was partitioned between 50 mL of sat. NaHCO3 and EtOAc.
The organic layer was washed with water and brine, dried over
Na2SO4, and concentrated in vacuo. The residue was purified by
column chromatography to afford compound 3 (1.43 g, 2.9 mmol)
1
as a yellow amorphous solid in 64% yield. H NMR (500 MHz,
DMSO-d6) δH 0.17 (s, 3H, SiCH3), 0.20 (s, 3H, SiCH3), 0.92 (s,
9H, CH3), 1.72 (d, J ) 1.0 Hz, 3H, T-CH3), 2.39 (ddd, J ) 1.4,
8.0 and 15.1 Hz, 1H, 2′-Ha), 2.75 (ddd, J ) 1.7, 7.1 and 15.1 Hz,
1H, 2′-Hb), 3.52 (d, J ) 12.6 Hz, 1H, 5′-Ha), 4.42 (d, J ) 15.3 Hz,
1H, 6′-Ha), 4.54 (d, J ) 15.3 Hz, 1H, 6′-Hb), 4.78 (d, J ) 12.5 Hz,
1H, 5′-Hb), 4.83 (dd, J ) 6.8 and 1.5 Hz, 1H, 3′-H), 6.09 (dd, J )
8.3 and 1.7 Hz, 1H, 1′-H), 7.20 (br s, 2H, NH2), 7.67 (d, J ) 1.1
Hz, 1H, T-6-H), 8.02 (s, 1H, A-2-H), 8.16 (s, 1H, A-8-H); 13C
NMR (500 MHz, DMSO-d6) δC -5.1, -4.8 (Si(CH3)2), 12.8 (T-
CH3), 17.7 (C(CH3)3), 25.6 (C(CH3)3), 43.3 (C-6′), 44.8 (C-2′), 74.0
(C-3′), 76.4 (C-5′), 88.2 (C-4′), 91.6 (C-1′), 117.1 (T-C-5), 118.2
(A-C-5), 138.7 (T-C-6), 141.3 (A-C-8), 149.8 (A-C-4); 152.5
(A-C-2), 155.9 (A-C-6), 156.2 (T-C-2), 170.9 (T-C-4); ESI
HRMS calcd for C22H32N7O4Si [M + H]+ 486.2285; found
486.2268.
4′-C-[(Adenin-9-yl)methyl]-3′-O-[(1,1-dimethylethyl)dimeth-
ylsilyl]-thymidine (4). To a solution of compound 3 (1.40 g, 2.88
mmol) in 45 mL of dioxane was added aqueous NaOH (2 N, 8.64
mL, 17.3 mmol) at room temperature, and the mixture was stirred
for 4 h. Then 8 mL of saturated NaHCO3 was added, and the
neutralized reaction mixture was concentrated in vacuo and
partitioned between 20 mL of H2O and 150 mL of EtOAc. The
organic layer was washed with water and brine, dried over Na2-
SO4, and concentrated in vacuo. The residue was purified by column
chromatography to afford compound 4 (0.67 g, 1.33 mmol) as a
yellow amorphous solid in 46% yield. 1H NMR (200 MHz, DMSO-
d6) δH 0.14 (s, 3H, SiCH3), 0.16 (s, 3H, SiCH3), 0.92 (s, 9H, CH3),
1.71 (s, 3H, T-CH3), 2.33-2.41 (m, 2H, 2′-H), 4.21 (d, J ) 16.0
Hz, 1H, 6′-Ha), 4.41 (d, J ) 16.0 Hz, 1H, 6′-Hb), 4.74 (t, J ) 6.0
Hz, 1H, 3′-H), 5.45 (br s, 1H, OH), 6.40 (t, J ) 6.1 Hz, 1H, 1′-H),
7.27 (br s, 2H, NH2), 7.63 (s, 1H, T-6-H), 8.09 (s, 1H, A-8-H),
8.15 (s, 1H, A-2-H), 11.32 (s, 1H, NH), 5′-H was obscured in a
water peak; 13C NMR (500 MHz, DMSO-d6) δC -4.8, -4.5 (Si-
(CH3)2), 12.5 (T-CH3), 18.0 (C(CH3)3), 26.0 (C(CH3)3), 43.9 (C-
6′), 61.6 (C-5′), 72.3 (C-3′), 83.5 (C-4′), 87.1 (C-1′), 109.9 (T-
C-5), 118.4 (A-C-5), 136.7 (T-C-6), 142.8 (A-C-8), 150.6 (A-
C-4); 150.9 (T-C-2), 153.1 (A-C-2), 156.2 (A-C-6), 164.4 (T-
C-4), C-2′ was hidden by solvent peaks; ESI HRMS calcd for
C22H34N7O5Si [M + H]+ 504.2391; found 504.2391.
Conclusion
Oligonucleotides have been synthesized containing bisheaded
furanose nucleosides in opposite strands and separated from each
other by one regular base pair. Oligonucleotides with a
bisheaded nucleotide may, likewise, be considered as a mimic
for base flipping, although the orientation of the base is different.
NMR analysis and molecular modeling have demonstrated that
the supplementary bases attached to a 4′-C-methyl group are
situated in the minor groove and may communicate with each
other mainly by hydrophobic interactions. Future research
activities will focus on the properties of the spacer between the
groove-oriented base and the sugar moiety to allow W-C
pairing of the extra nucleobases in the minor (or major) groove
of dsDNA.
Experimental Section
Chemical Synthesis. 5′-O-[(Trifluoromethyl)sulfonyl]-4′-C-
[(trifluoromethylsulfonyl]-oxymethyl]-3′-O-[(1,1-dimethylethyl)-
dimethylsilyl]-thymidine (2). To a solution of compound 1 (2.15
g, 5.56 mmol), which was prepared according to the reported
procedure,8 in 30 mL of anhydrous DCM was added 3.3 mL of
dried pyridine and trifluoromethanesulfonyl chloride (2.34 mL, 13.8
mmol) at 0 °C. The reaction mixture was stirred for 1 h and
partitioned between 20 mL of H2O and 200 mL of EtOAc. The
organic layer was washed with water and brine, dried over Na2-
SO4, and then concentrated in vacuo. The residue was purified by
3′,5′-O-Di[(1,1-dimethylethyl)dimethylsilyl]-4′-C-[(adenin-9-
yl)-methyl]-thymidine (5). To a solution of compound 4 (0.46 g,
5428 J. Org. Chem., Vol. 71, No. 15, 2006

Bisheaded DNA
0.92 mmol) in 25 mL of pyridine were added TBDMSCl (0.28 g,
1.84 mmol) and imidazole (0.16 g, 2.30 mmol) at room temperature.
The reaction mixture was stirred overnight. The solvent was
removed in vacuo and coevaporated twice with toluene. The residue
was partitioned between water and EtOAc. The organic layer was
washed with water and brine, dried over Na2SO4, and then
concentrated in vacuo. The residues were purified by column
chromatography. Compound 5 (0.33 g, 0.53 mmol) was obtained
as a colorless amorphous solid in 58% yield. 1H NMR (200 MHz,
CDCl3) δH 0.02 (s, 3H, SiCH3), 0.05 (s, 3H, SiCH3), 0.18 (s, 3H,
SiCH3), 0.19 (s, 3H, SiCH3), 0.90 (s, 9H, CH3), 0.97 (s, 9H, CH3),
1.88 (s, 3H, T-CH3), 2.19-2.27 (m, 1H, 2′-Ha), 2.32-2.48 (m, 1H,
2′-Hb), 3.29 (d, J ) 11 Hz, 1H, 5′-Ha), 3.82 (d, J ) 11 Hz, 1H,
5′-Hb), 4.32 (d, J ) 15 Hz, 1H, 6′-Ha), 4.50 (d, J ) 15 Hz, 1H,
6′-Hb), 4.74 (br d, J ) 8.2 Hz, 1H, 3′-H), 6.79 (dd, J ) 8.8 and 5.8
Hz, 1H, 1′-H), 7.23 (d, J ) 1.1 Hz, T-6-H), 8.23 (s, 1H, A-8-H),
8.36 (s, 1H, A-2-H); 13C NMR (200 MHz, CDCl3) δC -5.6, -5.5,
-5.1, -4.7 (Si(CH3)2), 12.3 (T-CH3), 18.0, 18.3 (C(CH3)3), 25.7,
25.8 (C(CH3)3), 40.5 (C-2′), 44.2 (C-6′), 65.4 (C-5′), 73.1 (C-3′),
81.8 (C-4′), 86.8 (C-1′), 111.3 (T-C-5), 118.0 (A-C-5), 135.3 (T-
C-6), 143.0 (A-C-8), 150.4 (A-C-4), 151.3 (T-C-2), 153.2 (A-
C-2), 155.6 (A-C-6), 165.6 (T-C-4); ESI HRMS calcd for
C28H48N7O5Si2 [M + H]+ 618.3255; found 618.3247.
70.3 (C-3′), 83.2 (C-4′), 86.9 (C-1′), 109.4 (T-C-5), 125.1 (A-
C-5), 128.6 (Bz-C), 132.6 (Bz-C), 133.6 (Bz-C), 136.2 (T-C-6),
145.9 (A-C-8), 150.3 (A-C-4), 150.6 (T-C-2), 151.6 (A-C-2),
153.4 (A-C-6), 163.9 (BzCO), 165.8 (T-C-4), C-2′ was hidden
by solvent peaks; ESI HRMS calcd for C23H24N7O6 [M + H]+
494.1788; found 494.1783.
5′-O-(Monomethoxytrityl)-4′-C-[(N6-benzoyladenin-9-yl)methyl]-
thymidine (8). To a solution of compound 7 (302 mg, 612 mmol)
in 20 mL of dried pyridine was added MMTrCl (245 mg, 796
mmol) at 0 °C. The reaction mixture was warmed to room
temperature and stirred overnight. Then 1 mL of methanol was
added to quench the reaction. The solvent was removed in vacuo
and coevaporated twice with toluene. The residue was partitioned
between water and EtOAc. The organic layer was washed with
water and brine, dried over Na2SO4, and then concentrated in vacuo.
The residue was purified by column chromatography to afford
compound 8 (310 mg, 496 mmol) as a colorless amorphous solid
1
in 81% yield. H NMR (200 MHz, DMSO-d6) δH 1.34 (s, 3H,
T-CH3), 3.70 (s, 3H, OCH3), 4.49 (br s, 2H, 6′-H), 4.87 (br s, 1H,
3′-H), 5.90 (br s, 1H, OH), 6.42 (t, J ) 6.4 Hz, 1H, 1′-H), 6.77 (d,
J ) 8.8 Hz, 2H, Ar H), 7.04 (d, J ) 8.8 Hz, 2H, Ar H), 7.23-
7.36(m, 10H, Ar H), 7.44 (s, 1H, T-6-H), 7.52-7.69 (m, 3H, Bz-
H), 8.04 (d, J ) 6.9 Hz, 2H, Bz-H), 8.22 (s, 1H, A-8-H), 8.59 (s,
1H, A-2-H), 11.13 (br s, 1H, NH), 11.33 (br s, 1H, NH), 2′-H was
obscured in solvent peaks. 5′-H was obscured in water peak at 3.35
ppm; 13C NMR (200 MHz, DMSO-d6) δC 11.7 (T-CH3), 44.4 (C-
6′), 55.1 (OCH3), 64.8 (C-5′), 70.8 (C-3′), 83.3 (C-4′), 86.2 (C-1′),
86.7 (C(Ar)3), 109.6 (T-C-5), 113.3 (ar-C), 124.6 (A-C-5), 127.1,
128.0, 128.6, 130.1, 132.5, 133.7, 134.3 (Ar-C), 135.9 (T-C-6),
143.7, 144.0 (Ar-C), 145.5 (A-C-8), 150.1 (A-C-4), 150.5 (T-
C-2), 151.3 (A-C-2), 153.0 (A-C-6), 158.4 (Ar-C), 163.9
(BzCO), 165.7 (T-C-4), C-2′ was hidden by solvent peaks; ESI
HRMS calcd for C43H40N7O7 [M + Na]+ 766.2989; found
766.2980.
3′,5′-O-Di[(1,1-dimethylethyl)dimethylsilyl]-4′-C-[(N6-benzoy-
ladenin-9-yl)methyl]-thymidine (6). To a solution of compound
B (320 mg, 0.52 mmol) in 20 mL of pyridine was added benzoyl
chloride (150 µL, 1.30 mmol) at 0 °C. The reaction mixture was
warmed to room temperature and stirred overnight. The solvent
was removed in vacuo and coevaporated twice with toluene. The
residue was partitioned between water and EtOAc. The organic
layer was washed with water and brine, dried over Na2SO4, and
then concentrated in vacuo. The oily residue was dissolved in
MeOH (20 mL), saturated with ammonia, and allowed to stand at
0 °C for half an hour. The mixture was concentrated, and the residue
was partitioned between water and EtOAc. The organic layer was
washed with water and brine, dried over Na2SO4, and then
concentrated in vacuo. The residue was purified by column
chromatography, affording compound 6 (353 mg, 0.49 mmol) as a
colorless amorphous solid in 94% yield. 1H NMR (200 MHz,
CDCl3) δH 0.02 (s, 3H, SiCH3), 0.04 (s, 3H, SiCH3), 0.19 (s, 3H,
SiCH3), 0.20 (s, 3H, SiCH3), 0.89 (s, 9H, CH3), 0.98 (s, 9H, CH3),
1.83 (s, 3H, T-CH3), 2.29-2.36 (m, 2H, 2′-H), 3.24 (d, J ) 11 Hz,
1H, 5′-Ha), 3.77 (d, J ) 11 Hz, 1H, 5′-Hb), 4.50 (br s, 2H, 6′-H),
4.69 (t, J ) 4.0 Hz, 1H, 3′-H), 6.63 (t, J ) 7.3 Hz, 1H, 1′-H),
7.45-7.59 (m, 4H, Bzm,p-H, T-6-H), 8.07 (d, J ) 6.6 Hz, 2H, Bzo-
H), 8.25 (s, 1H, A-8-H), 8.86 (s, 1H, A-2-H), 9.66 (br s, 1H, NH),
10.09 (br s, 1H, NH); 13C NMR (200 MHz, CDCl3) δC -5.6, -5.5,
-5.1, -4.7 (Si(CH3)2), 12.2 (T-CH3), 18.0, 18.2 (C(CH3)3), 25.7,
25.8 (C(CH3)3), 41.2 (C-2′), 45.0 (C-6′), 65.2 (C-5′), 73.8 (C-3′),
83.7 (C-4′), 87.6 (C-1′), 111.3 (T-C-5), 122.1 (A-C-5), 127.3,
128.3, 128.6, 132.6 (ar-C), 135.2 (T-C-6), 144.6 (A-C-8), 149.8
(A-C-4), 150.5 (T-C-2), 152.7 (A-C-2, A-C-6), 164.0 (Bz-CO),
165.3 (T-C-4); ESI HRMS calcd for C35H52N7O6Si2 [M + H]+
722.3518; found 722.3526.
5′-O-[(4-Methoxyphenyl)diphenylmethyl]-4′-C-[(N6-benzoy-
ladenin-9-yl)methyl]-3′-O-(P-â-cyanoethoxy-N,N-diisopropy-
laminophosphinyl)-thymidine (9). The monomethoxytritylated
derivative 8 (300 mg, 0.39 mmol) was dissolved in 10 mL of DCM
under argon, and diisopropylethylamine (DIEA, 204 µL, 1.17 mmol)
and 2-cyanoethyl N,N-diisopropylchlorophosphoramidite (CEPNCl,
175 µL, 0.78 mmol) were added. The mixture slowly became clear,
and overall the solution was stirred for 2 h, when TLC indicated
complete reaction. Water (2 mL) was added, and the solution was
stirred for 10 min and partitioned between CH2Cl2 (50 mL) and
aqueous NaHCO3 (30 mL). The organic phase was washed with
aqueous sodium chloride (2 × 30 mL), and the aqueous phases
were extracted with DCM (30 mL). Evaporation of the organic
solvent left a white foam that was purified twice on 40 g of silica
gel (hexane/acetone/TEA, 40:59:1) to afford the product as a foam
after several coevaporations with DCM. Dissolution in 2 mL of
DCM and precipitation in 50 mL of cold (-70 °C) hexane afforded
242 mg (0.25 mmol, 64%) of the title product 9 as a white powder.
Rf (hexane/acetone/TEA 40:59:1): 0.33. ESI-MS pos. calcd. for
C52H57N9O8P1 [M + H]+ 966.4067; found 966.4073; 31P NMR δ
ppm, (external ref ) H3PO4 capil.) 150.72, 151.02.
4′-C-[(Thymin-1-yl)methyl]-3′-O-[(1,1-dimethylethyl)dimeth-
ylsilyl]-thymidine (11). To a solution of thymine (3.35 g, 26.7
mmol) in 100 mL of DMF was added NaH (80%, 0.96 g, 32.0
mmol) at room temperature. The mixture was warmed to 70 °C
and kept stirring for 10 min. Once the reaction mixture was cooled
to 50 °C, a solution of compound 2 (1.73 g, 2.67 mmol) and a
catalytic amount of 18-crown-6 in 10 mL of DMF was added. After
being stirred for 4 h, the solvent was removed in vacuo. The residue
was partitioned between 50 mL of sat. NaHCO3 and 100 mL of
DCM. The organic layer was washed with water and brine, dried
over Na2SO4, and then concentrated in vacuo. The residue was used
directly in the next reaction, as compound 10 comigrated with
thymine and proved difficult to purify.
4′-C-[(N6-Benzoyladenin-9-yl)methyl]-thymidine (7). To a
solution of compound 6 (310 mg, 0.43 mmol) in THF was added
1 M TBAF (1.0 mL, 1.0 mmol) at room temperature. The reaction
mixture was kept stirring for 3-4 h and concentrated in vacuo.
The residue was purified by column chromatography to afford
compound 7 (120 mg, 0.24 mmol) as a colorless amorphous solid
1
in 57% yield. H NMR (200 MHz, DMSO-d6) δH 1.72 (s, 3H,
T-CH3), 2.34 (t, J ) 6.0 Hz, 2H, 2′-H), 3.18-3.27 (m, 1H, 5′-Ha),
3.44-3.53 (m, 1H, 5′-Hb), 4.41 (d, J ) 15 Hz 1H, 6′-Ha), 4.54 (d,
J ) 15 Hz, 1H, 6′-Hb), 4.61 (q, J ) 5.1 Hz, 3′-H), 5.30 (t, J ) 5.0
Hz, 1H, 5′-OH), 5.75 (d, J ) 4.8 Hz, 1H, 3′-OH), 6.41 (t, J ) 6.2
Hz, 1H, 1′-H), 7.52-7.65 (m, 4H, Bzm,p-H), 7.69 (br s, 1H, T-6-
H), 8.05 (d, J ) 7.0 Hz, 2H, Bzo-H), 8.42 (s, 1H, A-8-H), 8.74 (s,
1H, A-2-H), 11.15 (br s, 1H, NH), 11.27 (br s, 1H, NH); 13C NMR
(200 MHz, DMSO-d6) δC 12.2 (T-CH3), 43.8 (C-6′), 61.8 (C-5′),
J. Org. Chem, Vol. 71, No. 15, 2006 5429

Wu et al.
To a solution of this residue in 45 mL of dioxane was added
aqueous NaOH (2 N, 11 mL, 22 mmol) at room temperature, and
the mixture was stirred for 4 h. Then 15 mL of saturated NaHCO3
solution was added. The reaction mixture was concentrated in vacuo
to a small volume and partitioned between 20 mL of H2O and 100
mL of DCM. The organic layer was washed with water and brine,
dried over Na2SO4, and then concentrated in vacuo. A pure sample
as a colorless amorphous solid was obtained by preparative TLC.
1H NMR (500 MHz, DMSO-d6) δH 0.10 (s, 3H, SiCH3), 0.11 (s,
3H, SiCH3), 0.89 (s, 9H, C(CH3)3), 1.74 (s, 6H, T-CH3), 2.24-
2.29 (m, 1H, 2′-Ha), 2.33-2.39 (m, 1H, 2′-Hb), 3.37-3.46 (m, 2H,
5′-H), 3.79 (d, J ) 14.7 Hz, 1H, 6′-Ha), 3.92 (d, J ) 14.7 Hz, 1H,
6′-Hb), 4.68 (t, J ) 6.7 Hz, 1H, 3′-H), 5.23 (t, J ) 4.9 Hz, 1H,
OH), 6.26 (t, J ) 6.0 Hz, 1H, 1′-H), 7.47 (s, 1H, T*-6-H), 7.76 (s,
1H, T-6-H), 11.25 (br s, 2H, NH); 13C NMR (200 MHz, DMSO-
d6) δC -5.2, -4.8 (Si(CH3)2), 12.0*,12.3 (T-CH3), 17.7 (C(CH3)3),
25.6 (C(CH3)3), 46.9 (C-6′), 61.61 (C-5′), 71.4 (C-3′), 82.9 (C-1′),
87.2 (C-4′), 108.1*, 109.1 (T-C-5), 136.1, 142.6* (T-C-6), 150.4,
151.7* (T-C-2), 163.8, 164.1* (T-C-4), C-2′ was hidden by the
solvent peaks; ESI HRMS calcd for C22H34N4NaO7Si [M + Na]+
517.2094; found 517.2096.
4′-C-[(Thymin-1-yl)methyl]-thymidine (12). To a solution of
11 in THF was added 1 M TBAF (3.0 mL, 3.0 mmol) at room
temperature. The reaction mixture was stirred for 4 h and
concentrated in vacuo. The residue was purified by column
chromatography to afford compound 12 (121 mg, 0.32 mmol) as a
colorless amorphous solid in 12% yield from 2. 1H NMR (200 MHz,
DMSO-d6) δH 1.74 (s, 6H, T-CH3), 2.26 (t, J ) 6.4 Hz, 2H, 2′-H),
3.30-3.50 (m, 2H, 5′-H), 3.76 (d, J ) 15.0 Hz, 1H, 6′-Ha), 3.97
(d, J ) 15.0 Hz, 1H, 6′-Hb), 4.46-4.60 (m, 1H, 3′-H), 5.18 (t, J )
5.3 Hz, 1H, 5′-OH), 5.53 (d, J ) 4.4 Hz, 1H, 3′-OH), 6.27 (t, J )
6.0 Hz, 1H, 1′-H), 7.46 (app d, J ) 1.1 Hz, 1H, T-6-H), 7.79 (app
d, J ) 1.1 Hz, 1H, T-6-H), 11.27 (br s, 2H, NH); 13C NMR (200
MHz, DMSO-d6) δC 12.0, 12.3 (T-CH3), 47.1 (C-6′), 61.5 (C-5′),
70.1 (C-3′), 83.0 (C-1′), 87.5 (C-4′), 108.2, 109.2 (T-C-5), 136.3,
142.7 (T-C-6), 150.6, 152.0 (T-C-2), 163.9, 164.3 (T-C-4), C-2′
was hidden by solvent peaks; ESI HRMS calcd for C16H20N4NaO7
[M + Na]+ 403.1230; found 403.1230.
5′-O-(Monomethoxytrityl)-4′-C-[(thymin-1-yl)methyl]-thymi-
dine (13). To a solution of compound 12 (110 mg, 0.29 mmol) in
20 mL of dry pyridine was added MMTrCl (108 mg, 0.35 mmol)
at 0 °C. The reaction mixture was allowed to reach room
temperature and kept stirring overnight. Then 1 mL of methanol
was added. The solvent was removed in vacuo and coevaporated
twice with toluene. The residue was partitioned between water and
DCM. The organic layer was washed with water and brine, dried
over Na2SO4, and concentrated in vacuo. The residue was purified
by column chromatography to afford compound 13 as a colorless
amorphous solid (156 mg, 0.24 mmol) in 83% yield. 1H NMR (200
MHz, CDCl3) δH 1.54, 1.73 (two s, 6H, T-CH3), 2.29-2.53 (m,
2H, 2′-H), 3.06 (d, J ) 10.4 Hz, 1H, 5′-Ha), 3.34 (d, J ) 10.4 Hz,
1H, 5′-Hb), 3.78 (s 3H, OCH3), 4.05 (br s, (d, 2H, 6′-H), 4.23 (d,
J ) 5.8 Hz, 1H, OH), 4.60 (br s, 1H, 3′-H), 6.34 (t, J ) 6.6 Hz,
1H, 1′-H), 6.82 (d, J ) 8.8 Hz, 2H, Ar H), 7.06 (s, 1H, T-6-H),
7.24-7.48 (m, 13H, Ar H, T-6-H), 9.07 (s, 1H, NH), 9.19 (s, 1H,
NH); 13C NMR (200 MHz, CDCl3) δC 12.0 (T-CH3), 39.4 (C-2′),
48.7 (C-6′), 55.2 (OCH3), 66.0 (C-5′), 73.3 (C-3′), 84.5 (C-1′), 87.3
(C-4′ or C(Ar)3), 87.7 (C(Ar)3 or C-4′), 111.4, (T-C-5), 113.2 (T-
C-5), 113.3 (Ar-C), 127.2, 127.5, 127.9, 128.1, 128.3, 129.2, 130.4,
134.2 (ar-C), 135.7 (T-C-6), 141.6 (Ar-C), 143.3; 143.6 (Ar-
C), 150.2, 152.2 (T-C-2), 159.0, 163.6 (T-C-4); ESI HRMS calcd
for C36H36N4NaO8 [M + Na]+ 675.2431; found 675.2439.
5′-O-(Monomethoxytrityl)-4′-C-[(thymin-1-yl)methyl]-3′-O-(P-
â-cyanoethoxy-N,N-diisopropylaminophosphinyl)-thymidine (14).
The monomethoxytritylated derivative 13 (100 mg, 0.15 mmol) was
dissolved in 10 mL of DCM under argon, DIEA (80 µL, 0.45 mmol)
and CEPNCl (75 µL, 0.33 mmol) were added, and the solution
was stirred for 2 h, when TLC indicated incomplete reaction. An
additional quantity of 40 µL each of DIEA and CEPNCl were
added, and after being stirred for 1 h more, water (2 mL) was added.
The solution was stirred for 10 min and partitioned between DCM
(50 mL) and aqueous NaHCO3 (30 mL). The organic phase was
washed with aqueous sodium chloride (2 × 30 mL), and the
aqueous phases were back-extracted with CH2Cl2 (30 mL). Evapo-
ration of the organics left a light yellow foam that was flash purified
twice on 30 g of silica gel (hexane/acetone/TEA, 50:49:1) to afford
the product as a foam after several coevaporations with DCM.
Dissolution in 2 mL of DCM and precipitation in 50 mL of cold
(-70 °C) hexane afforded 77 mg (0.09 mmol, 60%) of the title
product 14 as a white powder. Rf (hexane/acetone/TEA, 40:59:1):
0.38; ESI-MS pos. calcd for C45H54N6O9P1 [M + H]+ 853.3690;
found 853.3715; 31P NMR δ ppm, (external ref ) H3PO4 capil.)
150.47, 150.95.
Oligonucleotide Synthesis. Oligonucleotide assembly was per-
formed on a DNA synthesizer using the phosphoramidite approach.
Standard DNA assembly protocols were used, either at the 0.2-
µmol scale or at the 10-µmol scale for the NMR oligonucleotides.
Coupling time for the modified bases was increased to 12 min using
0.07 M of the newly synthesized unnatural amidites, with ETT as
the activator substituting for tetrazole. The oligomers were depro-
tected and cleaved from the solid support by treatment with
concentrated aqueous ammonia (55 °C, 16 h). After gel filtration
with water as eluent, the crude was analyzed on an anion exchange
column, after which purification was achieved on a HR 10/10 with
the following gradient system (A ) 10 mM NaOH, pH 12.0, 0.1
M NaCl; B ) 10 mM NaOH, pH 12.0, 0.9 M NaCl; gradient used
depending on the oligo; flow rate 2 mL/min). The low-pressure
liquid chromatography system consisted of an intelligent pump, a
UV detector, and a recorder. The product-containing fraction was
desalted and lyophilized.
Oligonucleotides were characterized, and their purity was
checked by HPLC/MS on a capillary chromatograph. Columns of
150 mm × 0.3 mm length were used. Oligonucleotides were eluted
with a triethylammonium/1,1,1,3,3,3-hexafluoro-2-propanol/aceto-
nitrile solvent system. Flow rate was 5 µL/min. Electrospray spectra
were acquired on an orthogonal acceleration/time-of-flight mass
spectrometer in negative ion mode. Scan time used was 2 s. The
combined spectra from a chromatographic peak were deconvoluted
using the MaxEnt algorithm of the software. Theoretical oligo-
nucleotide masses were calculated using the monoisotopic element
masses.
Melting Temperatures. Oligomers were dissolved in 0.1 M
NaCl, 0.02 M potassium phosphate, pH 7.5, 0.1 mM EDTA. The
concentration was determined by measuring the absorbance in water
at 260 nm at 80 °C and assuming the cyclic nucleoside analogues
to have the same extinction coefficients in the denatured state as
the natural nucleosides with for TA
(ꢀ ) 23 500) and for TT (ꢀ )
17 000). The concentration in all experiments was 4 µM for each
strand unless otherwise stated. Melting curves were determined with
a UV spectrophotometer. Cuvettes were maintained at constant
temperature by means of water circulation through the cuvette
holder. The temperature of the solution was measured with a
thermistor directly immersed in the cuvette. Temperature control
and data acquisition were done automatically. The samples were
heated at a rate of 0.2 °C/min starting at 10 °C up to 80 °C and
cooling again at the same speed. Melting temperatures were
determined by plotting the first derivative of the absorbance versus
temperature curve and are the average of two runs. Up and down
curves in general showed identical Tm values.
NMR Experiments. NMR Sample Preparation. The duplexes
used for the NMR experiments [d(3′-GCGTTAAGCG-5′)-(5′-
CGCATTA CGC-3′)] and [d(3′-GCGTTAAGCG-5′)-(5′-CGCAT-
TCGC-3′)] were obtained by titrating (i) a solution of d(3′-
GCGTTAAGCG-5′) with the complementary sequence (5′-CGCATTA
CGC-3′) and (ii) a solution of d(3′-GCGTTAAGCG-5′) with the
complementary sequence (5′-CGCATTA CGC-3′). The degree of
complex formation was monitored by one-dimensional NMR spectra
of nonexchangeable base protons and anomeric protons. After
5430 J. Org. Chem., Vol. 71, No. 15, 2006

Bisheaded DNA
titration, the pD of the sample was adjusted to 7.2. The sample
was lyophilized and redissolved in 0.25 mL of D2O, resulting in a
concentration of 1.9 mM of the [d(3′-GCGTTAAGCG-5′)-(5′-
CGCATTA CGC-3′)] and 1.7 mM of [d(3′-GCGTTAAGCG-5′)-
(5′-CGCATTCGC-3′)]. The solutions were briefly heated to 80 °C
and slowly cooled to room temperature to promote duplex forma-
tion.
NMR Spectroscopy. Spectra were recorded on a spectrometer,
operating at 499.140 MHz. Unless stated otherwise, spectra were
recorded at 22 °C. Spectra were processed using the FELIX 97.00
software package running on a Silicon Graphics O2 R10000
workstation (IRIX version 6.3).
were constructed in a macromodel.22 Their geometry was optimized
in the Amber* force field.23 Then quatfit was used to fit the
modified nucleotides RTT on residue 15 and RTA on residue 6 of
the double helix.24
The energy of the system was minimized in Sander (Amber 6)
for 1000 steps, while maintaining NMR restraints (equilibrium H-H
distances of 5.0 Å for proton pairs obtained from observed NOE
contacts in the modified region, Figures 5 and 6). Examination of
the model showed that the extra bases had to be oriented in the
minor groove of the double helix.
The DNA was then solvated in a TIP3P water box.25 16 Na+
counterions were then added to get an electrostatic neutral system.
The water molecules and counterions were then allowed to relax
their positions while keeping the solute fixed. Molecular dynamics
simulations were then initiated with all restraints removed, with
periodic boundary conditions and using a cutoff distance of 9 Å
for the nonbonded interactions and the particle mesh Ewald method
for the summation of the Coulombic interactions.26 Simulation
temperature was 300 K and continued for 500 ps. During the last
50 ps, structures were collected. From those, one average structure
was calculated (shown in Figure 6a, detail shown in Figure 6b).
The NOE distances are marked in yellow dotted lines. Figures were
The TOCSY14 and NOESY15 spectra in D2O were recorded with
a sweep width of 4200 Hz in both dimensions. For the TOCSY
experiments, a clean MLEV1716 version was used, with a low power
90° pulse of 18.2 ms and the delay set to 47.3 ms. The total TOCSY
mixing time was set to 50 ms. The spectrum was acquired with 32
scans, 2048 data points in t2 and 512 FIDs in t1. The data were
apodized with a shifted sine bell square function in both dimensions
and processed to a 2K × 1K matrix. The NOESY experiments were
acquired with mixing times of 150 ms, 32 scans, 2048 data points
1
in t2 and 512 increments in t1. The natural abundance H, 13C
created by bobscript, molscript, and Raster3D.27-29
HSQC spectra were recorded with sensitivity enhancement and
gradient coherence selection17 using 72 scans and 256/512 complex
data points and 10 000/12 500 Hz spectral widths in t1 and t2,
respectively.
A
The proton-proton distances in the model are between H2(T6
)
T
and protons H1′(A14) 3.34 Å, HB71 (T15 ) 3.99 Å, H6′ (T15T) 4.34
T
Å, H1′ (T6 ) 4.28 Å, H5′ (T15T) 4.64 Å, confirming the NOE
measurements.
Molecular Modeling. A model of duplex 8 (Figure 6) was
created using the Amber software.18 Atomic electrostatic charges
to be used in the Amber software package were calculated from
the electrostatic potential at the 6-31G* level using the package
games19 and the two-stage RESP fitting procedure.20 The charges
of the atoms in the additional bases and atom C4′ were calculated
by the RESP procedure. All other charges were fixed to the values
as those of DNA in the Amber 94 topology files.21 The parameters
used in the Amber simulations are those from the parm99 dataset.20
An initial model of a nine-residue-long double-stranded DNA was
created using nucgen, which is part of the Amber suite of programs.
Modified T nucleotides with an additional adenine or thymine base
Acknowledgment. E.L. thanks the K. U. Leuven for the
postdoctoral fellow. K.N. is grateful to the FWO for his Ph.D.
fellowship. We thank C. Biernaux for the fine editorial help.
Financial support for this work was provided by the “Gecon-
certeerde Onderzoeksactie” of the Katholieke Universiteit Leu-
ven. Photograph for background of cover provided by Mrs.
Emily Stanfield with assistance from The Journal of Organic
Chemistry Editor-in-Chief Staff.
Supporting Information Available: General experimental
section and 1H NMR data of compounds 2, 3, 4, 5, 6, 7, 8, 11, 12,
13. This material is available free of charge via the Internet at
http://pubs.acs.org.
(14) Bax, A.; Davis, D. G. J. Magn. Reson. 1985, 65, 355-360.
(15) Jeener, J.; Meier, B. H.; Bachmann, P.; Ernst, R. R. J. Chem. Phys.
1979, 71, 4546-4553.
JO052194C
(16) Griesinger, C.; Otting, G.; Wuthrich, K.; Ernst, R. R. J. Am. Chem.
Soc. 1988, 110, 7870-7872.
(17) Schleucher, J.; Schwendinger, M.; Sattler, M.; Schmidt, P.; Schedletz-
ky, O.; Glaser, S. J.; Sorensen, O. W.; Griesinger, C. J. Biomol. NMR 1994,
4, 301-306.
(22) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput.
Chem. 1990, 11, 440-467.
(18) Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. R.;
Cheatham, T. E., III; DeBolt, S.; Ferguson, D.; Seibel, G.; Kollman, P.
Comput. Phys. Commun. 1995, 91, 1-41.
(23) Weiner, P.; Kollman, P. A.; Nguyen, D. T.; Case, D. A. J. Comput.
Chem. 1986, 7, 230-252.
(24) Heisterberg, D. J. Quatfit program in CCL software archives. The
Ohio Supercomputer Center: Columbus, OH.
(19) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A., Jr. J. Comput. Chem.
1993, 14, 1347-1363.
(25) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J.; Klein, M. L. J.
Chem. Phys. 1983, 79, 926-935.
(26) Cheatham, T. E., III; Miller, J. L.; Fox, T.; Darden, T. A.; Kollman,
P. A. J. Am. Chem. Soc. 1995, 117, 4193-4194.
(27) Esnouf, R. M. J. Mol. Graphics Model. 1997, 15, 132-134.
(28) Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 946-950.
(29) Merritt, E. A.; Bacon, D. J. Methods Enzymol. 1997, 277, 505-
524.
(20) Wang, J.; Cieplak, P.; Kollman, P. A. J. Comput. Chem. 2000, 21,
1049-1074.
(21) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman,
P. A. J. Am. Chem. Soc. 1995, 117, 5179-5197.
J. Org. Chem, Vol. 71, No. 15, 2006 5431