Bipyridyl- and biphenyl-DNA: A recognition motif based on interstrand aromatic stacking

Article abstract of DOI:10.1002/chem.200400858

The synthesis and incorporation into oligonucleotides of C-nucleosides containing the two aromatic, non-hydrogen-bonding nucleobase substitutes biphenyl (I) and bipyridyl (Y) are described. Their homo- and hetero-recognition properties in different sequential arrangements were then investigated via UV-melting curve analysis, gelmobility assays, CD- and NMR spectroscopy. An NMR analysis of a dodecamer duplex containing one biphenyl pair in the center, as well as CD data on duplexes with multiple insertions provide further evidence for the zipper-like interstrand stacking motif that we proposed earlier based on molecular modeling. UV-thermal melting experiments with duplexes containing one to up to seven I- or Y base pairs revealed a constant increase in T_m in the case of I and a constant decrease for Y. Mixed I/Y base pairs lead to stabilities in between the homoseries. Insertion of alternating I/abasic site- or Y/abasic site pairs strongly decreases the thermal stability of duplexes. Asymmetric distribution of I- or Y residues on either strand of the duplex were also investigated in this context. Duplexes with three natural base pairs at both ends and 50% of I pairs in the center are still readily formed, while duplexes with blunt ended I pairs tend to aggregate unspecifically. Duplexes with one natural overhang at the end of a I-I base pair tract can both aggregate or form ordered duplexes, depending on the nature of the natural bases in the overhang.

Full text of DOI:10.1002/chem.200400858

FULL PAPER
Bipyridyl- and Biphenyl-DNA:
A Recognition Motif Based on Interstrand Aromatic Stacking
Christine Brotschi, Gꢀrald Mathis, and Christian J. Leumann*[a]
Abstract: The synthesis and incorpora-
tion into oligonucleotides of C-nucleo-
sides containing the two aromatic, non-
hydrogen-bonding nucleobase substi-
tutes biphenyl (I) and bipyridyl (Y) are
described. Their homo- and hetero-rec-
ognition properties in different sequen-
tial arrangements were then investigat-
ed via UV-melting curve analysis, gel
mobility assays, CD- and NMR spec-
troscopy. An NMR analysis of a do-
decamer duplex containing one biphen-
yl pair in the center, as well as CD
data on duplexes with multiple inser-
tions provide further evidence for the
zipper-like interstrand stacking motif
that we proposed earlier based on mo-
lecular modeling. UV-thermal melting
experiments with duplexes containing
one to up to seven I- or Y base pairs
revealed a constant increase in Tm in
the case of I and a constant decrease
for Y. Mixed I/Y base pairs lead to sta-
bilities in between the homoseries. In-
sertion of alternating I/abasic site- or
Y/abasic site pairs strongly decreases
the thermal stability of duplexes.
Asymmetric distribution of I- or Y resi-
dues on either strand of the duplex
were also investigated in this context.
Duplexes with three natural base pairs
at both ends and 50% of I pairs in the
center are still readily formed, while
duplexes with blunt ended I pairs tend
to aggregate unspecifically. Duplexes
with one natural overhang at the end
of a I–I base pair tract can both aggre-
gate or form ordered duplexes, depend-
ing on the nature of the natural bases
in the overhang.
Keywords: bioorganic chemistry ·
DNA
recognition
·
hydrophobic effect · nucleosides
Introduction
tools for studying the fidelity and mechanistic aspects of
DNA-polymerase activity. It was shown that such isosters,
although destabilizing a DNA duplex, can code for each
other with high precision in DNA polymerase-based primer-
template extension reactions.[7–13]
The stabilities of such unnatural base pairs are strongly
dependent on the nature of the aromatic unit and vary
greatly from distinctly less stable to more stable compared
with a natural base pair. One of the first examples of a sta-
bilizing pair lacking hydrogen bonds was the pyrene/abasic
site pair.[6,12] Investigations towards the expansion of the ge-
netic alphabet from the Schultz and Romesberg group lead
to the production of a whole series of such stable, unnatural
self-pairs and cross-pairs.[14–22]
In an effort to decipher the role of interstrand stacking in-
teractions in DNA duplex stabilization with C-nucleosides
containing nonfunctionalized aromatic units (Figure 1) we
investigated bipyridyl self-pairs and found that two bipyridyl
C-nucleoside residues Y were able to recognize each other
within a DNA duplex with equal affinity as a G–C base
pair.[23] Based on molecular modeling we proposed a struc-
tural motif in which the distal pyridyl rings stack upon each
other with the interstrand stacking interaction as the main
driving force for the observed stability.
Hydrogen-bonding and stacking interactions between nucle-
obases are the major components of the noncovalent forces
that stabilize the DNA and RNA double helix.[1,2] The rela-
tive contribution of each to the stability has been a matter
of debate since the discovery of the double helix. About a
decade ago, Kool and co-workers started to investigate
stacking interactions in detail with oligodeoxynucleotide du-
plexes containing nonpolar nucleobase substitutes as dan-
gling ends.[3,4] Factors such as hydrophobicity (log P values),
polarizability, dipole moment, surface area and stacking
area were discussed as contributors to the observed en-
hanced thermodynamic stability.[4–6] Moreover, base substi-
tutes which mimic the shape of the natural bases but lack
their hydrogen-bonding capabilities (isosters) were used as
[a] Dipl.-Chem. C. Brotschi, Dipl.-Chem. G. Mathis, Prof. C. J. Leumann
Department of Chemistry and Biochemistry
University of Bern
Freiestrasse 3, 3012 Bern (Switzerland)
Fax : (+41)31-631-3422
E-mail: leumann@ioc.unibe.ch
Chem. Eur. J. 2005, 11, 1911 – 1923
DOI: 10.1002/chem.200400858
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1911

methylation in DNA.[25,26] However, its synthesis was differ-
ent from the one described here. The analogue 1c was ob-
tained along the same lines as described.[6] Our syntheses of
the C-nucleosides 1a and 1b followed established routes in
C-glycoside chemistry[27,28] (Scheme 1) and started with
2,3,5-tri-O-benzyl-d-ribonolactone 2[29] as the common deox-
yribose precursor to which the lithiated 4-bromo-2,2-biphen-
yl 3a or 5-iodo-2,2-bipyridine 3b[30,31] were added. While 3a
is commercially available, 3b had to be prepared in three
steps by stannylation of 2-bromopyridine 4 (!5),[32] fol-
lowed by a [Pd(PPh3)4] catalyzed Stille coupling with 2,5-di-
bromopyridine and a subsequent copper(i)-catalyzed halo-
gen-exchange reaction (Scheme 2). The halogen exchange
was necessary as, in our hands, the intermediate 5-bromo-
2,2-bipyridine 6[32] failed to react to the coupled product 7b
by the standard procedure.
Figure 1. Chemical structures of the hydrophobic nucleosides used in this
study.
Recently we extended our investigations also on pairs of
the simple biphenyl unit I and found an increase in duplex
stability of 3–4 K per I pair in duplexes containing up to
four of such base pairs. We proposed again a zipper-like in-
terstrand stacking motif of the distal biphenyl rings as the
most reasonable duplex structure.[24]
Lithiation of 3a or 3b followed by addition to lactone 2
resulted in the intermediate formation of the corresponding
hemiacetals, which upon reduction with Et3SiH/BF3·Et2O af-
Here we report on the synthesis and further investigation
of I and Y as hydrophobic base substitutes in various DNA
sequence contexts. We provide NMR-structural and CD-
spectroscopic evidence supporting the zipper-like interstrand
stacking motif and show scope and limitations of this recog-
nition motif as known so far.
Results
Scheme 2. Synthesis of 5-iodo-2,2-bipyridine. a) 4 (1 equiv), nBuLi
(1 equiv), Et2O, ꢀ788C then Me3SnCl (1 equiv) in THF, 93%; b) 2,5 di-
bromopyridine (1.1 equiv), [Pd(PPh3)4], m-xylene, reflux, 86%; c) CuI
(0.05 equiv), NaI (2 equiv), trans-N,N’-dimethyl-1,2-cyclohexanediamine
(0.1 equiv), dioxane, 1108C, 94%.
Synthesis of monomers: Of the hydrophobic nucleoside ana-
logues used in this work, 1a was already known and was in-
corporated into DNA as a tool to study enzymatic base
forded selectively the b-isomer-
ic C-glycosides 7a and 7b in
moderate but sufficient yield
(ca. 35%). The assignment of
the anomeric configuration in
7a and 7b was confirmed by
1H NMR NOE spectroscopy
(see Experimental Section).
After removal of the benzyl
protecting groups with BBr3 the
3’- and 5’-hydroxyl groups were
selectively protected with 1,3-
dichloro-1,1,3,3-tetraisopropyl-
disiloxane to give 8a and 8b,
again in moderate yield most
likely due to side reactions at
the benzylic, pseudo-anomeric
center C(1’). However, scram-
Scheme 1. a) 3a or 3b (1.0 equiv), nBuLi (1.0 equiv), THF, ꢀ788C, then
2
(0.9 equiv), ꢀ788C to RT;
bling of the configuration at
C(1’) during benzyl-deprotec-
tion could be excluded again by
1H NMR NOE evidence at the
stage of the tritylated com-
b) Et3SiH, (5 equiv), CH2Cl2, ꢀ788C to RT; c) BBr3 (3.4 equiv), CH2Cl2, ꢀ788C; d) 1,3-dichloro-1,1,3,3-tetrai-
sopropyldisiloxane (1.0 equiv), pyridine, RT; e) 1,1’-thiocarbonyldiimidazole (1.2 equiv), CH3CN, RT;
f) Bu3SnH (1.6 equiv), AIBN (1.5 equiv), toluene, 808C; g) NEt3·3HF (10.0 equiv), THF, RT; h) 4,4’-dimethoxy-
trityl (DMT) chloride (1.2 equiv), pyridine, RT; i) iPr2NEt (3 equiv), [(iPr2N)(NCCH2CH2O)P]Cl (1.5 equiv),
THF, RT. AIBN=a,a’-azobisisobutyronitrile.
1912
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
pounds 11a and the TIPS-protected intermediate 8b. Deox-
ygenation of the 2’-hydroxy group in 8a and 8b by means of
a Barton–McCombie reduction lead to 10a and 10b. Desily-
lation with (HF)·NEt3 gave the unprotected deoxynucleo-
side analogues 1a and 1b that were subsequently converted
into the corresponding phosphoramidite building blocks 12a
and 12b via tritylation (4,4’-dimethoxytrityl chloride) fol-
lowed by phosphitylation in typical yields for these transfor-
mations.
In the case of I all melting profiles show a single, highly
cooperative transition. For melting experiments with the
base Y 1 mm EDTA was added to the buffer in order to sup-
press metal chelation by the bipyridyl ligand. Also in these
cases only single transitions were observed. Examples of
representative melting curves are depicted in Figure 2. The
steeper base lines before and after the transition in melting
experiments with bipyridyl oligonucleotides were found to
be due to the EDTA in the buffer.
Synthesis of oligonucleotides: Oligonucleotides were pre-
pared according to standard protocols for automated DNA
synthesis on a 1 mmol scale in the trityl-off mode. The cou-
pling time for the modified phosphoramidites was extended
to 6–10 min, and ethyl thio-1H-tetrazole was used as activa-
tor. Typical coupling yields for the modified building blocks,
as deduced from the trityl assay, were in the range of 98%.
After assembly, the oligonucleotides were detached from
the solid support and deprotected in concentrated aqueous
ammonia (12–18 h at 558C) and were purified by HPLC. All
oligonucleotides were routinely characterized by ESIꢀ mass
spectrometry. The synthesis and analytical data for all modi-
fied oligonucleotides used in this study are summarized in
Table 6 (see Experimental Section).
Figure 2. UV-melting curves (260 nm) of duplexes containing I–I pairs
~
We then examined the structural and thermal melting
properties of duplexes by UV melting curve analysis, gel
mobility experiments, CD and NMR spectroscopy.
&
*
(
, n=3), Y–Y pairs ( , n=3) and I–Y mixed pairs ( , n=3).
The differences in Tm upon incorporation of a growing
Pairing properties of duplexes containing I or Y base pairs
in opposing positions (“stretched backbone” series): First,
we examined the stability of non self-complementary oligo-
nucleotide duplexes containing one or multiple aromatic
base pairs (I or Y) in the center of the sequence. The se-
quences were designed to determine not only homo base
pair but also hetero base pair formation. The thermal stabil-
ities of the corresponding duplexes were determined by UV-
melting curve analysis, and the corresponding Tm data are
summarized in Table 1.
number of I-, Y- or mixed base pairs are illustrated in
Figure 3. Incorporation of one I base pair leads to a drop in
Tm by 2.5 K relative to the unmodified duplex while addi-
tional consecutive I base pairs lead to an increase in Tm of
initially 4.4 K per base pair to 0.7 K for the incorporation of
the seventh base pair. On the other hand incorporation of
one Y base pair in the given sequence leads to a duplex that
is by 3 K more stable than the corresponding unmodified
duplex. However, contrary to the I series, any additional Y
base pair decreases duplex stability. The mixed, I–Y series
produces duplexes with almost invariant Tm relative to the
number of aromatic base pairs, with a tendency to lower Tm
for n=6. Thus the duplex stability of the mixed series lies in
between that of the two homoaromatic series.
Table 1. Duplex sequence information and a schematic representation of
the interstrand-stacking model, as well as corresponding Tm values.
[a,d]
[b,d]
5’-GATGAC(X)nGCTAG
3’-CTACTG(Z)nCGATC
Tm
[8C]
Tm
[8C]
Tm[a,d] [8C]
X=Y
Z=I
n
X=I
Z=I
X=Y
Z=Y
0
1
2
3
4
5
6
7
45.0
48.0
42.9
40.7
39.4
37.5
35.0
n.d.[c]
42.5
46.9
49.9
53.2
55.5
56.2
57.0
45.7
45.3
46.0
46.0
45.2
42.5
n.d.[c]
[a] c=1.2 mm duplex, 10 mm NaH2PO4, 0.15m NaCl, pH 7.0. [b] c=1.2 mm
duplex, see Table 2, 10 mm NaH2PO4, 0.15m NaCl,1 mm EDTA, pH 7.0.
[c] Not determined. [d] Estimated error in Tm =ꢁ0.58C.
Figure 3. Graphical representation of the stabilities of duplexes from
~
&
*
Table 1. I–I ( , n=1–7), Y–Y ( , n=1–6) and I–Y ( , n=1–6).
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1913

C. J. Leumann et al.
The higher stability of the duplex containing one Y base
pair compared to that with one I base pair (DTm =5.5 K)
seems to be in inverted order to the tendency of duplex sta-
bilization for the two aromatic units. This observation be-
comes plausible if one considers the stacking interactions of
the aromatic residues with the next natural base pair. Given
the fact that the bipyridyl system is intrinsically planar, it is
expected to perturb its nearest neighbor natural base pair
less than a biphenyl unit, which is intrinsically nonplanar.
NMR analysis of a duplex containing two opposing I units:
In order to get preliminary structural information we mea-
1
sured H NMR spectra of the duplex containing one I base
pair (Table 1, n=1) at different temperatures. The proton
signals arising from the biphenylic units in the duplex can
easily be identified by their chemical shift (6.5–7.0 ppm),
lying in between the signals of the nonexchangeable nucleo-
base protons and the anomeric sugar protons. Upon thermal
denaturation of the duplex the biphenyl signals shift in a co-
operative way towards lower field and end up at 7.1–
7.3 ppm in the single stranded state (Figure 4, shaded area).
This shift implies considerable loss of ring current effects
upon denaturation which is in agreement with the biphenyl
units stacking on each other in the duplex. Chemical shift
arguments have recently also been used as a criterion of in-
tercalation in a duplex containing a pyrene pair by others.[33]
The Tm that can be deduced from the cooperative shift of
the biphenyl protons versus T is at about 558C and thus
about 13 K higher than that measured by UV-melting
curves. This difference is expected and can be explained by
the 1000-fold higher oligonucleotide concentration in the
NMR sample. Thus the highly cooperative deshielding of
the biphenyl protons with a Tm in the expected range fully
supports the interstrand stacking model. A detailed structur-
al analysis of this duplex by 2D-NMR techniques is current-
ly underway and will be reported soon.
Figure 5. CD spectra of selected duplexes from Table 1. Top: base I, n=
0,1,3,5,6,7; c=3.6 mm in 10 mm NaH2PO4, 0.15m NaCl, pH 7, T=208C;
bottom: base Y, n=0,1,3,5. c=3.6 mm in 10 mm NaH2PO4, 0.15m NaCl,
1 mm EDTA, pH 7, T=208C.
modified duplex. The CD traces for I duplexes are com-
pared in Figure 5a, and those of the Y duplexes in Figure 5b.
In the case of the I duplexes, a shift of the negative maxi-
mum towards 250 nm with increasing numbers of I units is
observed. At the same time the positive maximum shifts
from 275 nm in the unmodified
CD spectra: A CD-spectroscopic investigation of selected
duplexes from Table 1 containing an increasing number of I-
or Y base pairs was performed and compared with the un-
duplex to about 270 nm and
grows in intensity as a function
of the number of I base pairs.
This is due to the contribution
of the biphenyl chromophores,
which have their maximum ab-
sorbance at about 250 nm. The
continuous shift of the CD
spectrum upon increasing the
number of I base pairs without
a sudden major change in its
shape is indicative for highly or-
dered helical structures and
disproves hydrophobic collapse
Figure 4. Temperature dependent 1H NMR spectra (500 MHz) of the DNA duplex 5’d(GATGACIGCTAG)-
d(CTAGCIGTCATC) (c=1 mm, 10 mm NaH2PO4, 150 mm NaCl, pH 7, 100% D2O).
of the central biphenylic core
part.
1914
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
In the case of the Y duplexes a similar picture emerges as
n is increased from 1 to 5 (Figure 5b). The minimum elliptic-
ity at about 255 nm in the unmodified duplex is shifted to-
wards shorter wavelengths (about 240 nm) as the number of
Y residues is increased. At the same time the maximum el-
lipticity shifts from about 275 to about 295 nm and gains in-
tensity. This again is most likely due to the bipyridyl chro-
mophore becoming increasingly CD active. Interestingly an
isodichroic point arises at 284 nm. From this we conclude
again that formation of consecutive Y base pairs occurs in a
highly ordered manner and leads to duplexes of one struc-
tural family, as in the case of the I base pairs. Thus also the
CD spectra are in support of an interstrand stacking model.
ing arrangements of I–H and Y–H pairs (H=abasic site). In
these systems interstrand stacking of the aromatic units is
still possible without necessarily invoking a change of the
backbone conformation relative to that of an unconstrained
B-DNA duplex (“relaxed backbone” series, Table 3).
Within this structural series we first investigated duplexes
containing two H–H, T–H and G–H pairs (Table 3, en-
tries 1–3). The Tm data clearly show an increase in the order
of the H–H duplex (198C) to the T–H duplex (27.58C) to
the G–H duplex (32.68C). This remarkable increase in Tm
goes parallel with the increasing potential of the bases to
overlap, and thus is in agreement with the interstrand stack-
ing model also in the “relaxed backbone” series. Within the
same structural motif the nonpolar aromatic I–H (35.78C),
Y–H (39.58C) and P–H (43.38C) pairs lead to duplexes with
considerably higher Tm values. The Tm values for the mixed,
I–H/Y–H, duplexes (entries 7 and 8) are lying in between
the Tm values for the I–H (entry 4) or Y–H (entry 5) substi-
tuted duplex. Thus, the order of stability of the mixed I- and
Y duplexes follows exactly that of the “nonrelaxed” back-
bone series (Table 1).
Melting experiments of duplexes containing four such
base pairs (entries 9–11) show a dramatic decrease of duplex
stabilities at low salt conc. (150 mm NaCl) up to an extent
where no distinct Tm value can be determined. Therefore
the Tm values were also measured at high salt (1m NaCl).
Interestingly the Y–H and I–H duplexes (entries 9 and 10)
led to higher Tm values than the P–H duplex (entry 11). This
is in agreement with earlier data on duplexes carrying up to
four P–H pairs that resulted in strongly reduced or no helix
formation, respectively.[6] The failure of duplex formation
with multiple P units may be due to competing hydrophobic
aggregation of these residues in the corresponding single
strands.
We also investigated duplexes where the Y–H and I–H
base pairs are in a nonalternating sequence context (Table 3,
entries 12 and 13). A comparison of the Tm values of the
corresponding alternating duplexes (entries 5 and 6) shows
that in the Y–H case the alternating arrangement is favored
by 3.2 K. In the I–H case there is no difference in Tm regard-
less whether the pairs are arranged in an alternating or non-
alternating manner. Generally, higher Tm values were ex-
pected in the alternating relative to the nonalternating cases
due to energetic differences arising from interstrand versus
intrastrand stacking. However, this is only encountered in
the Y–H case. Surprisingly it does not happen in the I–H
case. The reason for it is yet unknown, although we note
that differences in the preferred orientation of the biarylic
axis in the units I relative to Y could account for it. In any
case more structural data are necessary for a more detailed
interpretation of this fact.
Permutational analysis of the thermal stability of duplexes
with two consecutive I and Y base pairs: For the duplexes
with two consecutive aromatic units we evaluated the stabil-
ity of all eight possible arrangements of I and Y (Table 2,
entries 1–8). The Tm values of the six mixed duplexes are
lying in between the Tm values of the fully Y-substituted
duplex (42.98C) and the fully I-substituted duplex (46.98C).
The Tm values for the six mixed cases vary by 3.2 K.
Table 2. Permutational analysis of the thermal stability of pairs of I and
Y in a two-base system.
[a,c]
Duplex
Tm [8C]
5’-GATGACYYGCTAG
3’-CTACTGYYCGATC
5’-GATGACIIGCTAG
3’-CTACTGIICGATC
5’-GATGACIIGCTAG
3’-CTACTGYYCGATC
5’-GATGACYYGCTAG
3’-CTACTGIICGATC
5’-GATGACYIGCTAG
3’-CTACTGYICGATC
5’-GATGACYIGCTAG
3’-CTACTGIYCGATC
5’-GATGACYYGCTAG
3’-CTACTGYICGATC
5’-GATGACYYGCTAG
3’-CTACTGIYCGATC
1
2
3
4
5
6
7
8
42.9
46.9[b]
45.7
45.2
44.8
43.4
46.6
43.7
[a] c=1.2 mm, 10 mm NaH2PO4, 0.15m NaCl, 1 mm EDTA, pH 7.0.
[b] Same buffer without EDTA. [c] Estimated error in Tm =ꢁ0.58C.
The relative differences in duplex stability as a function of
the nature of the aromatic units is most expressed in the ho-
moaromatic systems and to a lesser extend in the mixed
series. Nevertheless there are distinct sequence effects which
eventually may establish a recognition code that is entirely
decoupled from hydrogen-bonding interactions and relies
only on differential stacking interactions, or possibly differ-
ential solvation.
CD experiments of selected duplexes (Table 3, entries 9–
11) are overlaid in Figure 6. No distinct deviations from a B-
DNA conformation are observed.
Pairing properties of duplexes containing I- or Y-abasic site
pairs in an alternating fashion (“relaxed backbone” series):
In contrast to I–I or Y–Y base pairs that upon intercalation
inevitably lead to a sugar phosphate backbone that is ex-
tended along the helical axis, we also investigated alternat-
Asymmetric distribution of I- and Y base pairs in the
duplex: Unpaired bases as occurring in hairpin–loop struc-
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1915

C. J. Leumann et al.
Table 3. Thermal stabilities of hydrophobic pairs in the “relaxed backbone” structural alignment, determined
by UV-melting curves; I (1a), Y (1b), P (1c), H (1d), G and T=deoxyguanosine and thymidine, respectively
of I units in both strands. In
system A (Table 4) we investi-
gated single bulge situations (n/
n+1, n=1–6) with both I- and
[a,c]
[b,c]
duplex
Tm [8C]
Tm [8C]
5’-GATGACHHGCTAG
3’-CTACTGHHCGATC
5’-GATGACTHGCTAG
3’-CTACTGHTCGATC
5’-GATGACGHGCTAG
3’-CTACTGHGCGATC
5’-GATGACIHGCTAG
3’-CTACTGHICGATC
5’-GATGACYHGCTAG
3’-CTACTGHYCGATC
5’-GATGACPHGCTAG
3’-CTACTGHPCGATC
5’-GATGACIHGCTAG
3’-CTACTGHYCGATC
5’-GATGACYHGCTAG
3’-CTACTGHICGATC
1
2
3
4
5
6
7
8
19.0
27.5
32.6
35.7
39.5
43.3
36.7
36.8
Y
residues. In system
B
(Table 4) we looked into du-
plexes with variable bulge size
(2/n+2, n=1–5) in the case of
the biphenylic unit I. The Tm
values for system A and B are
summarized in Table 4 and
graphically
Figure 7.
represented
in
In the I series, single bulges
within an increasing number of
I pairs (system A) go along
with an increase in the Tm of
the duplexes much in the same
way as their symmetric equiva-
lents (Table 1). The analogous
negative trend in Tm is observed
in the Y series. Interestingly,
asymmetric variation of the
bulge size (system B) in the
case of I is accompanied with
an almost invariant Tm which is
in between that of the symmet-
ric duplexes with n=2 and 3
(Table 1). Only with bulge sizes
larger than n=4 (Table 4,
system B, entries 2–6 and 2–7),
the duplexes become thermally
less stable. The CD spectra of
the duplexes of system A and B
are not significantly different
from that of the corresponding
symmetric sequences (data not
shown).
5’-GATGAHIHICTAG
3’-CTACTIHIHGATC
5’-GATGAHYHYCTAG
3’-CTACTYHYHGATC
5’-GATGAHPHPCTAG
3’-CTACTPHPHGATC
9
~5
~10
~5
20.0
23.0
18.0
10
11
5’-GATGACYYGCTAG
3’-CTACTGHHCGATC
5’-GATGACIIGCTAG
3’-CTACTGHHCGATC
12
13
36.3
35.6
[a] c=1.2 mm, 10 mm NaH2PO4, 0.15m NaCl, pH 7.0. [b] c=1.2 mm, 10 mm NaH2PO4, 1.0m NaCl, pH 7.0. [c] Es-
timated error in Tm =ꢁ0.58C.
tures or in bulged structures are common natural structural
motifs. We became interested in the question of how the
biarylic units I and Y behave in bulge positions and there-
fore investigated duplexes with an asymmetric distribution
Sequences with terminal I stretches or without natural base
pairs: To explore whether natural base pairs flanking the I
pairs are necessary in order to maintain discrete duplex
Figure 7. Graphical representation of the thermal stabilities (Tm) vs the
^
)
number of I residues (n) of duplexes in the asymmetric systems A (
*
Figure 6. CD spectra (T=208C) of selected duplexes from Table 3.
and B ( ) from Table 4.
1916
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
Table 4. Thermal stabilities of duplexes containing an asymmetric distribution of hydrophobic pairs. System
A: n/n+1; System B: 2/2+n; I (1a), Y (1b).
two strands with themselves or
with each other. Aggregate for-
System A
n/n+1
X=I
Duplex
X=Y
Tm
mation occurs also with
a
[b,d]
Tm[a,d] [8C]
[8C]
strand containing a 3’-natural
overhang (entry 3) or with one
of the single strands (entry 1)
alone. By contrast, the duplex
5’-GATGACXGCTAG
3’-CTACTGXXCGATC
5’-GATGACXXGCTAG
3’-CTACTGXXXCGATC
5’-GATGACXXXGCTAG
3’-CTACTGXXXXCGATC
1/2
2/3
3/4
4/5
5/6
6/7
42.9
47.5
51.0
54.2
56.2
57.2
X=I
43.7
41.4
containing
a 5’-natural over-
39.6
hang (entry 4) shows a clear
and highly sigmoidal transition
in the UV-melting curve (Tm
51.38C) and a single duplex
band in the gel (Figure 8,
entry 4). This Tm is only 5.7 K
lower than that of the duplex
containing natural base pairs at
both ends (Table 5, entry 5) and
is 42.4 K higher than that of the
natural hexamer duplex consist-
ing of the first six base pairs
(Tm =8.98C).[34] From these ex-
periments we conclude that
stable I pairs can exist also in
duplexes with natural base pairs
on one end and a 5’-overhang
of natural bases on the other
end, while duplexes with blunt-
ended I pairs or a 3’-overhang
tend to aggregate. The reason
of the differential behavior of
duplexes with 5’- or 3’-overhang
is unknown at present, but is
expected to be due to stacking
of the overhang on the neigh-
boring I pair. In another gel
mobility experiment we investi-
gated duplexes with shortened
ends of natural base pairs
Figure 8, entries 6–8). In this
case it becomes clear that three
natural base pairs at both ends
suffice to produce a discrete
duplex structure. Its Tm could
not be determined due to the
observed low hyperchromicity.
In a next step we wanted to
5’-GATGACXXXXGCTAG
3’-CTACTGXXXXXCGATC
5’-GATGACXXXXXGCTAG
3’-CTACTGXXXXXXCGATC
5’-GATGACXXXXXXGCTAG
3’-CTACTGXXXXXXXCGATC
38.4
n.d.[c]
n.d.[c]
System B
2/n+2
Duplex
[a]
Tm
5’-GATGACXXGCTAG
3’-CTACTGXXXCGATC
5’-GATGACXXGCTAG
3’-CTACTGXXXXCGATC
5’-GATGACXXGCTAG
2/3
2/4
2/5
2/6
2/7
47.5
47.9
47.1
44.1
42.4
3’-CTACTGXXXXXCGATC
5’-GATGACXXGCTAG
3’-CTACTGXXXXXXCGATC
5’-GATGACXXGCTAG
3’-CTACTGXXXXXXXCGATC
[a] c=1.2 mm, 10 mm NaH2PO4, 0.15m NaCl, pH 7.0. [b] c=1.2 mm, 10 mm NaH2PO4, 0.15m NaCl, 1 mm EDTA,
pH 7.0. [c] Not determined. [d] Estimated error in Tm =ꢁ0.58C.
Table 5. I-modified oligonucleotide single strands and duplexes used in the gel retardation experiments, the
corresponding Tm data from UV-melting curves and discernible secondary structure formation.
[a,d]
Entry
1
Sequence
Tm
[8C]
Secondary structure
aggregate
5’-GATGACIIIIIII
5’-GATGACIIIIIII
3’-CTACTGIIIIIII
mmp[b]
mmp[b]
2
3
4
aggregate
aggregate
duplex
5’-GATGACIIIIIIIGCTAG
3’-CTACTGIIIIIII
mmp[b]
51.3
5’-GATGACIIIIIII
3’-CTACTGIIIIIIICGATC
5’-GATGACIIIIIIIGCTAG
3’-CTACTGIIIIIIICGATC
5’-GATIIIIIITAG
5
6
7
57.0
–
duplex
single strand
duplex
5’-GATIIIIIITAG
3’-CTAIIIIIIATC
3’-CTAIIIIIIATC
5’-TIIIIIIIIIIIIT
n.d.
8
9
10
–
n.d.
n.d.
single strand
aggregate
single strand
5’-TTTTTTTTIIIIIIIIIIIIT
5’-TTTTTTTTIIIIIIIIIIIIT
3’-TIIIIIIIIIIIIT
11
n.d.
mixed aggregate
explore whether
a
pure
I
[a] c=1.2 mm, 10 mm NaH2PO4, 0.15m NaCl, pH 7.0. [b] Multiple overlaying melting processes were observed
[c] n.d.=no Tm detectable due to low hyperchromicity. [d] Estimated error in Tm =ꢁ0.58C.
duplex can exist without the
help of flanking natural base
pairs. Due to the expected low
hyperchromicity upon melting
structures we prepared and analyzed the duplexes listed in
Table 5. A 13-mer duplex containing seven I pairs and six
natural base pairs at one end (Table 5, entry 2) failed to
show a sigmoidal transition in the UV-melting curve. A gel
mobility experiment (Figure 8, lane 2) showed multiple
bands reminiscent of less-defined aggregate formation of the
of a fully biphenylic system, we adopted an experimental ap-
proach developed by Switzer et al.[35] We prepared two
single strands containing 12 I residues each, that were
tagged with one and eight thymidine residues at their 5’-
end, respectively, resulting in different oligonucleotide
lengths (Table 5, entries 9 and 10). Assuming that the flank-
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1917

C. J. Leumann et al.
diester backbone. It is notable that this zipper corresponds
to an intercalator/base-pair ratio of 1:1; this indicates that
even further extension along the helical axis is tolerated by
the backbone. It is unclear at present if and where the limi-
tations in length of this recognition motif lie. The fact that
the “relaxed backbone” series with Y–H and I–H pairs are
in all cases less stable than the “stretched backbone” series
with I–I and Y–Y pairs may well have its origin in differen-
ces in the entropy of duplex formation, the former series
having more degrees of conformational freedom than the
latter series.
Figure 8. 20% Polyacrylamide gel electrophoresis of single strands and
mixtures corresponding to entries 1–11 from Table 5). Bands were visual-
ized by UV light (260 nm). Entries correspond to Table 5.
An interesting question is also, whether the RNA back-
bone, which is intrinsically more compact and stiffer, will
tolerate such a zipper-like intercalation motif or not. This
would certainly have implications on research in the area of
novel base pairs for biotechnological applications.[14–22]
ing T residues do not alter strand affinities, in a 1:1 mixture
of both strands one would expect three bands in the case of
statistic duplex formation. However, no discrete band of a
single strand was found for the shorter oligonucleotide
(Figure 8, lane 9), while a sharp band with the expected mo-
bility of a single strand was observed for the longer oligonu-
cleotide (entry 10). The mixture of both showed the disap-
pearance of the band of the longer oligonucleotide without
the appearance of discrete new bands (entry 11) eventually
indicating cross-pairing.
Fully modified duplexes: An obvious question is, whether
fully modified duplexes built only from I- or Y base pairs
exist and are stable under standard conditions. We found
that defined DNA duplex formation from two single strands
containing six I residues and three natural base pairs at
either end readily takes place (Figure 8, lane 7). Further-
more, we know that natural base pairs on one end of the
duplex can (but must not) be sufficient for defined duplex
formation (Figure 8, lane 4). On the other hand, oligonu-
cleotides that have no matching natural base pairs, or that
show blunt ended I pairs (Figure 8, lane 1, 2, 9–11) either
tend to aggregation or do not pair at all. Therefore a few
natural base pairs at one end of the duplex seem to be re-
quired to align two strands in a defined structural register
and to prevent undefined aggregation.
Discussion
I- versus Y pairs: The increasing thermal stability of duplex-
es with an increasing number of I pairs contrasts findings in
the Y series where decreasing thermal stability with increas-
ing number of bipyridyl residues is observed. Obvious fac-
tors responsible for this behavior are differences in stacking
or solvation energies, or differences in the preferred orienta-
tion of the biarylic axis. In general, Kool et al.[4] found that
hydrophobic effects are more important in stabilizing stack-
ing than other effects (electrostatic effects, dispersion
forces). However, the natural DNA bases are found to be
less dependent on hydrophobic effects than the more non-
polar compounds. We exclude differences in the polarizabili-
ty (am) of the two aromatic systems as the major factor, as
calculations revealed very similar am values (18.64 ꢁ3 for Y,
both syn and anti isomer, 20.15 ꢁ3 for I). More likely, differ-
ences in solvation of the edges of the aromatic units in the
major or minor groove in the duplex state, or structural dif-
ferences around the biphenylic axis (intrinsically planar in
the case of Y and intrinsically nonplanar in the case of I)
are responsible for the stability differences. The latter could
also explain why a single I base pair is less stable compared
with a single Y base pair, if its stacking interaction with the
nearest neighbor natural base pair is taken into account.
Conclusion
With the structural and biophysical data contained in this
communication we lend further evidence for the existence
of the interstrand-stacking structural model of non-hydrogen
bonding aromatics as base substitutes that we first proposed
a couple of years ago.[23,24] These findings underline the
more and more recognized importance of interstrand stack-
ing interactions for nucleic acid double-helix stability. In ad-
dition it demonstrates the structural dynamics of the DNA
backbone towards an elongation along the helical axis.
Besides representing an ideal scaffold for the study of the
energetics of stacking interactions it could also be a structur-
al motif of interest in nanotechnology or molecular electron-
ics. In this context we are currently interested in the follow-
ing questions: i) how do stacking energies depend on the
electronic nature of the aromatic units; ii) does there exist a
recognition code based entirely on differential stacking in-
teractions; iii) can donor and acceptor substituted biphenyls
form extended charge transfer complexes when alternatingly
stacked within the center of a helix; and iv) can electrons or
charges be efficiently transported across the zipper much in
the same way as in natural DNA, and can this be kinetically
Relaxed versus stretched backbone: It is long known that
ethidium bromide can intercalate up to an extent of one
ethidium per 2.5 base pairs in DNA–polymer duplexes.[1]
This illustrates the flexibility of the DNA backbone which
allows for considerable breathing along the helical axis. It is
thus not surprising that a zipper-motif as described here in
the “stretched backbone” series is tolerated by the phospho-
1918
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
94%) was obtained after FC (hexane/EtOAc/TEA 8:2:2%) as a white
solid. Rf =0.30 (hexane/EtOAc/TEA 8:2:2%); m.p. 1048C; 1H NMR
(300 MHz, CDCl3, 258C): d =8.86 (d, 3J(H,H)=1.5 Hz, 1H; C6H), 8.65
(dd, 3J(H,H)=0.8, 4.8 Hz, 1H; C6H), 8.36 (d, 3J(H,H)=8.1 Hz, 1H;
C3H), 8.20 (d, 3J(H,H)=7.8 Hz, 1H; C3H), 8.10 (dd, 3J(H,H)=2.2,
8.5 Hz, 1H; C4H), 7.80 (td, 3J(H,H)=1.8, 7.7 Hz, 1H; C4H), 7.31 (ddd,
3J(H,H)=1.1, 4.8, 5.9 Hz, 1H; C5H); 13C NMR (75 MHz, CDCl3, 258C):
d =155.18 (s), 155.10 (d), 154.83 (s), 149.12 (d), 145.11 (d), 136.98 (d),
124.02 (d), 122.74 (d), 120.88 (d), 93.83 (s); MS (70 eV, EI): m/z (%): 282
(100) [M]+, 155 (57), 128 (25), 78 (30); IR (KBr): n˜ =1586, 1573, 1558,
and thermodynamically tuned via electronic variation
(chemical substitution) of the biphenyl units.
Experimental Section
General: Reactions were performed under argon in distilled, anhydrous
solvents. All chemicals were reagent grade from Fluka or Aldrich.
1H NMR (300 MHz, 500 MHz) spectra were recorded on a Bruker AC-
300 or Bruker DRX-500 spectrometer; the chemical shifts d/ppm were
referenced to residual undeuterated solvent ([D]chloroform)=7.27,
[D]methanol=3.35); coupling constants J in Hz. 13C NMR (75 MHz)
were recorded on a Bruker AC300; the chemical shifts d/ppm were refer-
enced to residual undeuterated solvent ([D]chloroform=77.00,
[D]methanol=49.3). Carbon multiplicity (s,d,t,q) from DEPT spec-
tra.31P NMR spectra (162 MHz) were recorded on a Bruker DRX-400
spectrometer; the chemical shift d/ppm was referenced to 85% H3PO4 as
external standard. LSIMS and EI mass spectra were recorded on a Auto
Speq Q VG at 70 eV, ESI-MS mass spectra on a Fisons Instrument VG
Platform. For TLC, pre-coated plates SIL-G UV254 (Macherey Nagel)
have been used and visualized by UV and/or dipping into a solution of
Ce(SO4)2 (10.5 g), phosphormolybdic acid (21 g), H2SO4 (60 mL), and
H2O (900 mL). Flash chromatography (FC) was performed with silica gel
60 (230–400 mesh). Abbreviations: EtOAc: ethyl acetate; TEA: triethyl-
amine; TIPDS: 1,3-dichloro-1,1,3,3-tetraisopropyl disiloxane; DMT: 4,4’-
dimethoxytrityl.
1540, 1455, 1435, 1358, 998, 791, 633 cmꢀ1
Compound 7a: nBuLi (1.6m in hexane, 9.3 mL, 14.9 mmol) was added
dropwise at ꢀ788C to solution of 4-bromo-biphenyl (3a, 3.5 g,
15.0 mmol) in dry THF (130 mL). After 1 h at ꢀ788C, (6.16 g,
.
a
2
14.71 mmol) in dry THF (20 mL) was added. The mixture was stirred for
3 h at ꢀ788C and allowed to warm up to RT over night. Sat. NaHCO3
(300 mL) was added and the mixture was extracted with Et2O (3ꢂ
250 mL). The organic layer was washed with H2O (100 mL), dried
(MgSO4) and concentrated in vacuo. The yellow residue was dissolved in
dry CH2Cl2 (40 mL) and cooled to ꢀ788C. Et3SiH (11.7 mL, 73.6 mmol)
and BF3·OEt2 (9.3 mL, 73.6 mmol) were added dropwise. The reaction
mixture was allowed to warm to RT over night, quenched with HCl (1m,
30 mL) and stirred for 30 min at RT. The mixture was neutralized with
2% NaOH and extracted with EtOAc (4ꢂ150 mL). The organic layer
was washed with brine (50 mL), dried (MgSO4) and concentrated in
vacuo. Compound 7a (2.93 g, 35%) was obtained after FC (hexane/
EtOAc 9:1) as a slightly yellow solid. Rf =0.15 (hexane/EtOAc 9:1); m.p.
1
3
80–878C; H NMR (300 MHz, CDCl3, 258C): d =7.65–7.63 (d, J(H,H)=
2.7 Hz, 2H; ArH), 7.60–7.54 (m, 4H; ArH), 7.51–7.46 (m, 3H; ArH),
7.44–7.23 (m, 15H; ArH), 5.15 (d, 3J(H,H)=6.6 Hz, 1H; C1’H), 4.69–
4.56 (m, 6H; CH2), 4.44 (m, 1H; C4’H), 4.11 (t, 3J(H,H)=4.8 Hz, 1H;
C3’H), 3.93 (t, 3J(H,H)=5.9 Hz, 1H; C2’H), 3.73 (ddd, 2,3J(H,H)=4.0,
10.7, 14.7 Hz, 2H; C5’H); 1H NMR NOE (400 MHz, CDCl3, 258C): d =
5.15 (C1’H) ! 7.61 (ArH; 8.8%), 4.44 (C4’H; 3.0%), 3.93 (C2’H; 1.7%);
4.44 (C4’H) ! 5.15 (C1’H; 3.4%), 4.11 (C3’H; 2.3%), 3.73 (C5’H;
2.7%); 4.11 (C3’H) ! 3.93 (C2’H, 7.6%); 13C NMR (75 MHz, CDCl3,
258C): d =140.98 (s, ArC), 140.57 (s, ArC), 139.47 (s, ArC), 138.17,
137.96, 137.80 (3s, ArC), 128.73, 128.37, 128.28, 128.07, 127.75, 127.64,
127.60, 127.19, 127.06, 127.02, 126.72 (11d, ArC), 83.71 (d, C2’), 82.38 (d,
C1’), 81.78 (d, C4’), 77.54 (d, C3’), 73.47, 72.25, 71.95 (3t, CH2), 70.47 (t,
C5’); MS (70 eV, EI): m/z (%): 556 (0.05) [M]+, 465 (5), 448 (3), 357
(10), 153 (12), 107 (8), 91 (100), 77 (13).
Compound 5: A solution of nBuLi (1.6m in hexane, 62.7 mL, 104 mmol)
was diluted with dry Et2O (45 mL).
A solution of 4 (10.2 mL,
103.7 mmol) in dry Et2O (100 mL) was added at ꢀ788C. The mixture was
stirred for 1 h at ꢀ788C. A solution of Me3SnCl (20.0 g, 104 mmol) in dry
THF (100 mL) was then added dropwise at ꢀ788C. The cooling bath was
removed and the reaction mixture was allowed to warm to RT. The color
changed from brown to green and finally to yellow. The reaction was
quenched with sat. NH4Cl and the organic layer was washed with H2O,
brine, dried (MgSO4) and evaporated. Compound 5 (21.2 g, 93%) was
obtained by distillation under reduced pressure as a colorless liquid.
1H NMR (300 MHz, CDCl3, 258C): d = 8.73 (d, 3J(H,H)=4.4 Hz, 1H),
7.54–7.42 (m, 2H), 7.15–7.11 (m, 1H), 0.34 (s, 9H; CH3); 13C NMR
(75 MHz, CDCl3, 258C): d =173.51 (s), 150.51 (d), 133.51 (d), 131.58 (d),
122.24 (d), ꢀ9.50 (q); MS (70 eV, EI): m/z (%): 243/241 (25/20) [M]+,
228/226 (100/75), 198/196 (77/42), 135/133 (53/41).
Compound 7b: This compound was prepared as described for 7a, from
3b (6.0 g, 0.021 mol), nBuLi (13.3 mL) in THF (280 mL). Lactone 2
(8.35 g, 0.0199 mol) in THF (70 mL) was added. After extraction the
yellow residue was dissolved in CH2Cl2 (54.5 mL), Et3SiH (15.81 mL,
0.0995 mol) and BF3·OEt2 (12.5 mL, 0.0995 mol) were added. Compound
7b (3.51 g, 32%) was obtained after FC (hexane/EtOAc 1:1 followed by
a second column hexane/EtOAc/TEA 8:2:2%) as a slightly yellow oil.
Rf =0.21 (hexane/EtOAc/TEA 9:1:2%); 1H NMR (500 MHz, CDCl3,
258C): d = 8.73 (d, 3J(H,H)=2.7 Hz, 1H; ArH), 8.71 (ddd, 3J(H,H)=
1.2, 2.4, 6.1 Hz, 1H; ArH), 8.44 (d, 3J(H,H)=9.9 Hz, 1H; ArH), 8.35 (d,
3J(H,H)=10.2 Hz, 1H; ArH), 7.87–7.83 (m, 2H; ArH), 7.40–7.30 (m,
10H; ArH), 7.26–7.24 (m, 3H; ArH), 7.20–7.18 (m, 2H; ArH), 7.18 (d,
3J(H,H)=3.0 Hz, 1H; ArH), 5.10 (d, 3J(H,H)=9.3 Hz, 1H; C1’H), 4.63–
4.51 (m, 5H; CH2), 4.44 (d, 3J(H,H)=15.0 Hz, 1H; CH2), 4.40 (dd,
3J(H,H)=4.7, 8.9 Hz, 1H; C4’H), 4.05 (dd, 3J(H,H)=4.0, 6.4 Hz, 1H;
C3’H), 3.85 (dd, 3J(H,H)=6.6, 9.4 Hz, 1H; C2’H), 3.68 (dd, 2,3J(H,H)=
5.0, 12.9 Hz, 1H; C5’H), 3.62 (dd, 2,3J(H,H)=5.0, 12.9 Hz, 1H; C5’H);
Compound 6: 2,5-Dibromopyridine (36.9 g, 0.156 mol) was added to a so-
lution of 2-trimethylstannylpyridine (5, 34.5 g, 0.141 mol) in m-xylene
(300 mL), and the reaction mixture was degassed by gently bubbling
argon through the solution for 1 h. Then, [Pd(PPh3)4] (1.64 g, 1 mol%)
was added and heated to 1208C under stirring. After 12 h the mixture
was cooled and poured into 2m NaOH. The phases were separated, and
the aqueous layer was extracted with toluene (2ꢂ1000 mL). The com-
bined organic phases were dried (MgSO4) and evaporated under reduced
pressure. Compound 6 (28.53 g, 86%) was obtained after FC (hexane/
EtOAc/TEA 8:2:2%) as a white solid. Rf =0.30; 1H NMR (300 MHz,
CDCl3, 258C): d = 8.73 (d, 3J(H,H)=2.5 Hz, 1H), 8.67 (dt, 3J(H,H)=
0.8, 3.2 Hz; 1H), 8.38 (d, 3J(H,H)=7.9 Hz, 1H), 8.32 (d, 3J(H,H)=
8.5 Hz, 1H), 7.94 (dd, 3J(H,H)=2.3, 8.5 Hz, 1H) 7.82 (td, 3J(H,H)=1.7,
7.5 Hz, 1H), 7.32 (ddd, 3J(H,H)=1.5, 4.7, 7.4 Hz, 1H); 13C NMR
(75 MHz, CDCl3, 258C): d =155.04 (s), 154.42 (s), 150.18 (d), 149.08 (d),
139.49 (d), 137.14 (d), 124.00 (d), 122.36 (d), 121.17 (s), 121.02 (d); MS
(70 eV, EI): m/z (%): 234/236 (100/100) [M]+, 155 (76), 128 (82), 78 (39).
1H NMR NOE (500 MHz, CDCl3, 258C): d
=
5.10 (C1’H) ! 4.44
Compound 3b: A Schlenk tube was charged with compound 6 (8 g,
0.034 mol), CuI (0.323 g, 0.0017 mol) and NaI (10.2 g, 0.068 mol), briefly
evacuated and flushed with argon. Racemic trans-N,N’-dimethyl-1,2-cy-
clohexanediamine[28] (0.544 mL, 0.0034 mol) and dioxane (34 mL) were
added. The Schlenk tube was sealed and the reaction mixture stirred at
1108C for 70 h. The brown suspension was allowed to reach RT, was di-
luted with 30% aqueous ammonia (180 mL), poured into water (500 mL)
and extracted with CH2Cl2 (3ꢂ200 mL). The combined organic layers
were dried (MgSO4) and concentrated in vacuo. Compound 3b (9.02 g,
(C4’H; 3.0%); 4.05 (C3’H) ! 3.85 (C2’H; 11%); 3.85 (C2’H) ! 4.05
(C3’H, 10%); 13C NMR (75 MHz, CDCl3, 258C): d =155.94 (s, ArC),
155.54 (s, ArC), 149.08 (d, ArC), 147.47 (d, ArC), 137.89 (s, ArC), 137.72
(s, ArC), 137.36 (s, ArC), 136.87 (d, ArC), 136.04 (s, ArC), 134.75 (d,
ArC), 128.38–127.58 (15d, ArC), 123.59 (d, ArC), 121.06 (d, ArC), 120.68
(d, ArC), 83.80 (d, C2’), 82.30 (d, C1’), 80.07 (d, C4’), 77.42 (d, C3’),
73.49, 72.50, 71.92 (3t, CH2), 70.38 (t, C5’); LSIMS: m/z (%): 559 (100)
[M+H]+, 185 (37), 147 (32), 136 (45).
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1919

C. J. Leumann et al.
Compound 8a:
A
solution of 7a (2.23 g, 4.0 mmol) in dry CH2Cl2
1H; C3’), 4.33–4.28 (m, 1H; C5’H), 4.17–4.10 (m, 2H; C4’H, C5’H),
1.17–0.94 (m, 28H; iPr); 13C NMR (75 MHz, CDCl3, 258C): d =141.2,
140.68, 137.69, 136.81, 136.80 (5s, ArC), 130.81, 128.80, 127.44, 127.30,
127.12, 126.89, 126.43, 124.08, 118.10 (9d, ArC, ImC), 87.64 (d, C2’),
82.89 (d, C1’), 82.12 (d, C4’), 69.21 (d, C3’), 60.72 (t, C5’), 17.48, 17.36,
17.34, 17.27, 17.08, 16.98, 16.90 (7q, iPr), 13.37, 13.05, 12.89, 12.67 (4d,
CH-iPr); LSIMS: m/z (%): 639 (38) [M+H]+, 511 (10), 467 (8), 261 (75),
233 (100).
(40 mL) was cooled to ꢀ788C and treated with BBr3 (1m in CH2Cl2,
14.0 mL, 14.0 mmol). After 4 h at ꢀ788C the reaction mixture was
quenched with MeOH (45 mL) and allowed to warm up to RT over
night. The solvent was evaporated; the residue dissolved in MeOH
(200 mL) and washed with hexane (3ꢂ40 mL). The aqueous layer was
evaporated and residual water was removed by coevaporation from pyri-
dine. The crude product (1.14 g) was dissolved in dry pyridine (44 mL)
and 1,3-dichloro-1,1,3,3-tetraiso-propyldisiloxane (1.25 mL, 4.0 mmol)
was added dropwise at 08C. After stirring for 5 h at RT the solvent was
evaporated. The residue was dissolved in saturated aqueous NaHCO3
(80 mL) and extracted with EtOAc (4ꢂ40 mL). The organic layer was
dried (MgSO4) and concentrated in vacuo.
Compound 9b: This compound was prepared as described for 9a, from
8b (139 mg, 0.262 mmol), 1,1’-thiocarbonyl diimidazole (56 mg,
0.31 mmol) in acetonitrile (1.2 mL). The residue was dissolved in EtOAc
(50 mL) and washed with saturated aqueous NaHCO3 (3ꢂ40 mL). Com-
pound 9b (120 mg, 71%) was obtained after FC (EtOAc/hexane 4:1) as
1
a white foam. Rf =0.25 (EtOAc/hexane 3:1); H NMR (300 MHz, CDCl3,
Compound 8a (761 mg, 36% over
2 steps) was obtained after FC
258C): d =8.89 (d, 3J(H,H)=1.8 Hz, 1H; ArH), 8.69 (d, 3J(H,H)=
(hexane/EtOAc 9:1) as a slightly yellow oil. Rf =0.24 (hexane/EtOAc
8:2); 1H NMR (300 MHz, CDCl3, 258C): d =7.62–7.58 (m, 4H; ArH),
7.53–7.43 (m, 4H; ArH), 7.38–7.33 (m, 1H; ArH), 4.91 (d, 3J(H,H)=
3.7 Hz, 1H; C1’H), 4.43 (t, 3J(H,H)=6.6 Hz, 1H; C3’H), 4.15–4.13 (m,
4.0 Hz, 1H; ArH), 8.45 (m, 1H; ImH), 8.43–8.42 (m, 2H; ArH), 8.07
3
3
(dd, J(H,H)=1.8, 8.0 Hz, 1H; ArH), 7.84 (dd, J(H,H)=1.8, 7.7 Hz, 1H;
ArH), 7.71 (m, 1H; ImH), 7.35–7.31 (m, 1H; ArH), 7.09 (s, 1H; ImH),
5.78 (d, 3J(H,H)=4.0 Hz, 1H; C2’H), 5.36 (s, 1H; C1’H), 4.67 (dd,
3J(H,H)=5.2, 8.8 Hz, 1H; C3’H), 4.29 (m, 1H; C5’H), 4.18–4.09 (m, 2H;
C4’H, C5’H), 1.11–0.92 (m, 28H; iPr); 13C NMR (75 MHz, CDCl3, 258C):
d=183.35 (s, C=S), 156.17, 155.72 (2s, ArC), 149.16 (d, ArC), 147.27 (d,
ArC), 136.87 (d, ImC), 136.77, 134.53, 134.29 (3d, ArC), 131.08 (d, ImC),
130.67, 123.79, 121.13 (3d, ArC), 118.04 (d, ImC), 87.08 (d, C2’), 82.17 (d,
C1’), 81.09 (d, C4’), 69.16 (d, C3’), 60.53 (t, C5’), 17.41, 17.30, 17.27,
17.22, 17.01, 16.92, 16.88, 16.84 (8q, iPr), 13.26, 12.97, 12.83, 12.61 (4d,
CH-iPr); LSIMS: m/z (%): 641 (10) [M+H]+, 531 (15), 261 (18), 235
(40), 278 (64), 252 (66), 215 (40), 181 (28), 160 (38), 105 (18).
3
2H; C5’H), 4.09–4.05 (m, 1H; C4’H), 4.02 (dd, J(H,H)=3.7, 5.9 Hz, 1H;
C2’H), 3.20–2.80 (br, 1H; HO), 1.14–1.04 (m, 28H; iPr); 13C NMR
(75 MHz, CDCl3, 258C): d =140.91 (s, ArC), 140.60 (s, ArC), 139.11 (s,
ArC), 128.75 (d, ArC), 127.26 (s, ArC), 127.17 (d, ArC), 127.11 (d, ArC),
126.27 (d, ArC), 85.26 (d, C1’), 82.46 (d, C2’), 77.30 (d, C4’), 71.62 (d,
C3’), 62.45 (t, C5’), 17.51, 17.40, 17.38, 17.33, 17.15, 17.11, 17.00 (7q, iPr),
13.42, 13.20, 12.89, 12.65 (4d, CH-iPr); MS (70 eV, EI): m/z (%): 528
(0.8) [M]+, 485 (29), 467 (15), 455 (9), 395 (7), 235 (100), 205 (29), 152
(17), 77 (9).
Compound 8b: This compound was prepared as described for 8a, from
7b (552 mg, 0.99 mmol), BBr3 (1m in CH2Cl2, 3.45 mL, 3.45 mmol) in
CH2Cl2 (10 mL). The reaction was quenched with MeOH (12 mL), the
mixture evaporated and the residue dissolved in H2O (50 mL) and
washed with CH2Cl2 (1ꢂ20 mL). The aqueous layer was evaporated and
dried by coevaporation from pyridine. To the crude product (285 mg) 1,3-
dichloro-1,1,3,3-tetraisopropyldisiloxane (0.37 mL, 0.99 mmol) in dry pyr-
idine (11 mL) was added. Compound 8b (239 mg, 46% over two steps)
was obtained after FC (EtOAc/hexane 3:8) as a slightly yellow oil. Rf =
0.26 (EtOAc/hexane 3:8); 1H NMR (300 MHz, CDCl3, 258C): d =8.71
(d, 3J(H,H)=2.2 Hz, 1H; ArH), 8.70–8.68 (m, 1H; ArH), 8.43–8.38 (m,
2H; ArH), 7.90 (dd, J(H,H)=1.9, 8.1 Hz, 1H; ArH), 7.83 (dt, J(H,H)=
1.8, 7.7 Hz, 1H; ArH), 7.32 (ddd, 3J(H,H)=1.1, 4.8, 7.4 Hz, 1H; ArH),
4.92 (d, 3J(H,H)=3.7 Hz, 1H; C1’H), 4.40 (t, 3J(H,H)=7.8 Hz, 1H;
C3’H), 4.16–4.06 (m, 3H; C4’H, C5’H), 3.99 (m, 1H; C2’H), 3.06 (d,
3J(H,H)=4.1 Hz, 1H; (HO), 1.12–1.02 (m, 28H; iPr); 1H NMR NOE
(400 MHz, CDCl3, 258C): d =4.92 (C1’H) ! 8.70 (ArH; 6.4%), 7.90
(ArH; 2.7%), 4.06 (C4’H; 4.2%), 3.99 (C2’H; 2.6%); 4.40 (C3’H) ! 8.71
(ArH; 1.1%), 7.90 (ArH; 2.2%), 3.99 (C2’H; 10.7%); 3.99 (C2’H) !
8.70 (ArH; 2.2%), 7.91 (ArH, 1.5%), 4.91 (C1’H; 3.2%), 4.41 (C3’H;
10.0%); 4.06 (C4’H) ! 4.92 (C1’H; 3.6%), 4.40 (C3’H; 2.5%); 13C NMR
(101 MHz, CDCl3, 258C): d =155.73 (s, ArC), 155.42 (s, ArC), 148.95 (d,
ArC), 146.91 (d, ArC), 136.81 (d, ArC), 135.48 (s, ArC), 134.27 (d, ArC),
123.53 (d, ArC), 121.01 (d, ArC), 120.63 (d, ArC), 83.09 (d, C1’), 82.70
(d, C2’), 76.53 (d, C4’), 71.57 (d, C3’), 62.29 (t, C5’), 17.31, 17.20, 17.18,
17.14, 16.94, 16.91, 16.89, 16.78 (8q, iPr), 13.20, 13.01, 12.74, 12.46 (4d,
CH-iPr); LSIMS: m/z (%): 531 (100) [M+H]+, 199 (19), 185 (28), 133
(14).
Compound 10a: A solution of 9a (700 mg, 1.1 mmol) in toluene (7.2 mL)
was heated at 808C, and a solution of Bu3SnH (0.58 mL, 2.19 mmol) and
a,a’-azobisisobutyronitrile (34 mg, 0.21 mmol) in toluene (18 mL) was
added dropwise over a period of 2 h. The reaction mixture was stirred for
2 h at 808C and then allowed to cool to RT. The mixture was concentrat-
ed in vacuo. Compound 10a (438 mg, 78%) was obtained after FC
(hexane/EtOAc 15:0.3) as a colorless oil. Rf =0.61 (hexane/EtOAc 8:2);
1H NMR (300 MHz, CDCl3, 258C): d =7.62–7.58 (m, 4H; ArH), 7.48–
7.34 (m, 5H; ArH), 5.17 (t, 3J(H,H)=7.35 Hz, 1H; C1’H), 4.60 (m, 1H;
C3’H), 4.19 (m, 1H; C5’H), 3.96–3.90 (m, 2H; C4’H, C5’H), 2.48–2.40
(m, 1H; C2’H), 2.20–2.11 (m, 1H; C2’H), 1.14–1.07 (m, 28H; iPr);
13C NMR (75 MHz, CDCl3, 258C): d =141.09, 140.93, 140.44 (3s, ArC),
128.71, 127.19, 127.11, 127.06, 126.29 (5d, ArC), 86.46 (d, C4’), 78.84 (d,
C1’), 73.43 (d, C3’), 63.76 (t, C5’), 43.11 (t, C2’), 17.59, 17.46, 17.44, 17.39,
17.26, 17.11, 17.08, 16.99 (8q, iPr), 13.52, 13.40, 13.03, 12.57 (4d, CH-iPr);
MS (70 eV, EI): m/z (%): 512 (0.02) [M]+, 469 (86), 451 (19), 439 (5), 235
(100).
3
3
Compound 10b: This compound was prepared as described for 10a.
Bu3SnH (80 mL, 0.30 mmol) and a,a’-azobisisobutyronitrile (4.6 mg,
28 mol) in toluene (0.4 mL) were added to 9b (119 mg, 0.186 mmol) in
toluene (2 mL). Compound 10b (69 mg, 72%) was obtained after FC
(EtOAc/hexane/TEA 1:3:0.05) as a slightly yellow oil. Rf =0.36 (EtOAc/
hexane 1:3); 1H NMR (300 MHz, CDCl3, 258C): d =8.67 (d, 3J(H,H)=
4.0 Hz, 1H; ArH), 8.62 (d, 3J(H,H)=1.5 Hz, 1H; ArH), 8.39 (d,
3J(H,H)=4.8 Hz, 1H; ArH), 8.36 (d, 3J(H,H)=5.2 Hz, 1H; ArH), 7.84–
7.77 (m, 2H, ArH), 7.31–7.27 (m, 1H; ArH), 5.18 (t, 3J(H,H)=7.4 Hz,
1H; C1’H), 4.56 (m, 1H; C3’H), 4.15 (m, 1H; C5’H), 3.97–3.87 (m, 2H;
C4’H, C5’H) 2.45 (ddd, 2,3J(H,H)=4.8, 7.0, 12.5 Hz, 1H; C2’H), 2.15–2.03
(m, 1H; C2’H), 1.09–1.03 (m, 28H; iPr); 13C NMR (75 MHz, CDCl3,
Compound 9a: 1,1’-Thiocarbonyl diimidazole (300 mg, 1.7 mmol) was
added to
a solution of 8a (740 mg, 1.40 mmol) in dry acetonitrile
258C):
d =156.23, 155.72 (2s, ArC), 149.35, 147.33, 137.85, 137.05,
(6.1 mL). After stirring for 9 h at RT a second portion of 1,1’-thiocarbon-
yl diimidazole (150 mg, 0.85 mmol) was added and stirred for another
9 h. The suspension was concentrated in vacuo, the residue dissolved in
CH2Cl2 (100 mL) and washed with saturated aqueous NaHCO3 solution
(3ꢂ40 mL). The organic layer was dried (MgSO4) and concentrated in
vacuo. Compound 9a (751 mg, 84%) was obtained after FC (hexane/
EtOAc 7:3) as a white solid. Rf =0.41 (hexane/EtOAc 7:3); m.p. 112–
1168C; 1H NMR (300 MHz, CDCl3, 258C): d =8.54 (s, 1H; ImH), 7.76
(s, 1H; ImH) 7.64–7.59 (m, 6H; ArH), 7.46 (t, 3J(H,H)=7.35 Hz, 2H;
ArH), 7.39–7.35 (m, 1H; ArH), 7.16 (s, 1H; ImH), 5.84 (d, 3J(H,H)=
4.8 Hz, 1H; C2’H), 5.32 (s, 1H; C1’H), 4.71 (dd, 3J(H,H)=4.8, 8.8 Hz,
134.55, 123.81, 121.24, 120.96 (8d, ArC), 86.81 (d, C4’), 77.62 (d, C1’),
73.47 (d, C3’), 63.79 (t, C5’), 43.21 (t, C2’), 17.76, 17.64, 17.61, 17.56,
17.43, 17.27, 17.16, 16.06 (8q, iPr), 13.80, 13.69, 13.58, 13.24 (4d, CH-iPr);
LSIMS: m/z (%): 515 (100) [M+H]+, 183 (20).
Compound 1a: NEt3·3HF (1.0 mL, 6.3 mmol) was added dropwise to a
solution of 10a (323 mg, 0.63 mmol) in dry THF (32 mL). After stirring
for 15 h at RT the reaction mixture was concentrated in vacuo. Com-
pound 1a (167 mg, 98%) was obtained after FC (EtOAc/toluene/MeOH
5:5:0.3) as
a white solid. Rf =0.37 (EtOAc/toluene/MeOH 5:5:1);
1H NMR (300 MHz, [D]methanol, 258C): d=7.65–7.60 (m, 4H; ArH),
1920
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
7.51–7.33 (m, 5H; ArH), 5.21 (dd, 3J(H,H)=5.2, 10.7 Hz, 1H; C1’H),
4.40–4.38 (m, 1H; C3’H), 4.05–4.00 (m, 1H; C4’H), 3.75 (m, 2H; C5’H),
(NaSO4) and concentrated in vacuo. Compound 12a (284 mg, 89%) was
obtained after FC (toluene/EtOAc/TEA 8:2:0.2) as a white foam. Rf =
0.72 (toluene/EtOAc/TEA 8:2:0.2); 1H NMR (300 MHz, CDCl3, 258C): d
3
2.30–2.23 (ddd, J(H,H)=1.5, 5.5, 13.3 Hz, 1H; C2’H), 2.07–1.96 (m, 1H;
C2’H); 13C NMR (75 MHz, [D]methanol, 258C)
d
=142.57, 142.42,
=7.61–7.18 (m, 18H; ArH), 6.84 (d, J(H,H)=3.7 Hz, 2H; ArH), 6.81 (d,
3
142.06 (3s, ArC), 130.13, 128.59, 128.19, 127.97 (4d, ArC), 89.49 (d, C4’),
81.65 (d, C1’), 74.74 (d, C3’), 64.36 (t, C5’), 45.22 (t, C2’); MS (70 eV, EI):
m/z (%): 270 (53) [M]+, 252 (4), 234 (10), 180 (83), 167 (100), 152 (44),
115 (18), 77 (19); IR (KBr): n˜ =3394, 2927, 1486, 1084, 1071, 1045, 1008,
3J(H,H)=3.67 Hz, 2H; ArH), 5.22 (m, 1H; C1’H), 4.55 (dd, 3J(H,H)=
5.9, 10.7 Hz, 1H; C4’H), 4.27 (m, 1H; C3’H), 3.84–3.71 (m, 2H;
CH2CO), 3.79 (s, 6H; OCH3), 3.65–3.52 (m, 2H; CH-iPr), 3.41–3.23 (m,
2H; C5’H), 2.65 (t, 3J(H,H)=6.30 Hz, 1H; CH2CN), 2.50 (t, 3J(H,H)=
6.30, 1H; CH2CN), 2.45–2.39 (m, 1H; C2’H), 2.13 (m, 1H; C2’H), 1.22–
1.09 (m, 12H; CH3-iPr); 13C NMR (75 MHz, CDCl3, 258C): d =158.39,
144.87, 140.90, 140.87, 140.75, 140.72, 136.09, 136.03 (8s, ArC), 130.12,
130.09, 128.68, 128.26, 128.22, 127.72, 127.15, 127.02, 126.70, 126.66,
126.49 (11d, ArC), 177.48, 117.42 (2s, CN), 113.03 (d, ArC), 86.04 (s,
ArC), 85.98, 85.72, 85.64 (3d, C4’H), 80.08, 80.01 (2d, C1’), 76.34, 76.11,
75.92, 75.70 (4d, C3’), 64.18, 64.13 (2t, C5’), 58.42, 58.39, 58.17, 58.14 (4t,
CH2), 55.14, 55.13 (2q, OCH3), 43.25, 43.21 (2t, C2’), 43.08, 43.04 (2d,
CH-iPr), 24.64, 24.52, 24.46, 24.36 (4q, CH3-iPr), 20.34, 20.25, 20.19, 20.09
(4t, CH2CN); 31P NMR (202 MHz, CDCl3, 258C): d =149.30, 149.09;
HRMS, LSI-MS: m/z: calcd for C47H54N2O6P1: 773.3719; found: 773.3687
[M+H]+.
830, 764, 701 cmꢀ1
.
Compound 1b: This compound was prepared as described for 1a, from
10b (61 mg, 0.12 mmol) and NEt3·3HF (0.20 mL, 1.2 mmol) in THF
(7 mL). Compound 1b (31 mg, 95%) was obtained after FC (EtOAc/tol-
uene/MeOH/TEA 5:5:3:2%) as a slightly pink oil. Rf =0.48 (EtOAc/tolu-
ene/MeOH/TEA 5:5:3:2%); 1H NMR (300 MHz, [D]methanol, 258C): d
=8.72–8.69 (m, 2H; ArH), 8.33–8.31 (m, 2H; ArH), 8.06–7.95 (m, 2H;
ArH), 7.48 (m, 1H; ArH), 5.29 (dd, 3J(H,H)=5.5, 10.7 Hz, 1H; C1’H),
4.43–4.41 (m, 1H; C3’H), 4.00 (m, 1H; C4’H), 3.75 (d, 3J(H,H)=5.2 Hz,
2H; C5’H), 2.34 (dd, 3J(H,H)=5.5, 13.23 Hz, 1H; C2’H), 2.11–1.96 (m,
1H; C2’H); 13C NMR (75 MHz, [D]methanol, 258C): d =157.23, 156.67
(2s, ArC), 150.52, 148.67 (2d, ArC), 140.04 (s, ArC), 139.04, 136.77,
125.58, 122.93, 122.59 (5d, ArC), 89.77 (d, C4’), 79.41 (d, C1’), 74.69 (d,
C3’), 64.22 (t, C5’), 45.10 (t, C2’); LSIMS: m/z (%): 273 (15) [M+H]+,
155 (100), 135 (18), 119 (89).
Compound 12b: This compound was prepared as described for 12a, from
11b (30 mg, 0.06 mmol), N,N-diisopropylethylamine (30 mL, 0.17 mmol)
and 2-cyanoethyl diisopropylchlorophosphoramidite (20 mL, 0.09 mmol)
in dry THF (1.6 mL). Compound 12b (28 mg, 69%) was obtained after
FC (EtOAc/hexane/TEA 1:1:2%) as a slightly yellow foam. Rf =0.45,
0.53 (EtOAc/hexane/TEA 1:1:2%); 1H NMR (300 MHz, CDCl3, 258C):
Compound 11a: 4,4’-Dimethoxytrityl chloride (220 mg, 0.65 mmol) was
added in three portions to a solution of 1a (146 mg, 0.54 mmol) in dry
pyridine (2.2 mL). After 6 h, toluene (2 mL) was added and the reaction
mixture was evaporated. The residue was dissolved in CH2Cl2 (30 mL)
and washed with saturated aqueous NaHCO3 (3ꢂ15 mL), dried (MgSO4)
and concentrated in vacuo. Compound 11a (240 mg, 85%) was obtained
d
= 8.69–8.67 (m, 2H; ArH), 8.42–8.36 (m, 2H; ArH), 7.89 (dt,
3J(H,H)=2.9, 8.1 Hz, 1H; ArH), 7.82 (td, 3J(H,H)=1.5, 7.7 Hz, 1H;
ArH), 7.48 (m, 2H; ArH), 7.38–7.20 (m, 10H; ArH), 6.85–6.81 (m, 4H;
ArH), 5.27 (dd, 3J(H,H)=4.8, 10.3 Hz, 1H; C1’H), 4.60–4.55 (m, 1H;
C4’), 4.29 (m, 1H; C(3’H), 3.81–3.64 (m, 8H; CH2CO, OCH3), 3.62–3.59
(m, 2H; CH-iPr), 3.35–3.27 (m, 2H; C5’H), 2.63 (t, 3J(H,H)=6.3, 1H;
CH2CN), 2.51–2.45 (m, 1H; CH2CN) 2.11–2.05 (m, 1H; C2’H), 1.75–1.65
(m, 1H; C2’H), 1.22–1.10 (m, 12H; CH3-iPr); 13C NMR (75 MHz, CDCl3,
258C): d =158.48 (s, ArC), 156.04, 155.57 (2s, ArC), 149.15, 147.32 (2d,
ArC), 144.75 (2s, ArC), 137.28 (s, ArC), 136.85 (d, ArC), 136.00, 135.97
(2s, ArC), 134.58 (d, ArC), 130.1, 128.13, 128.25, 128.2, 127.80, 126.78
(5d, ArC), 123.61, 121.04, 120.76 (3d, ArC), 117.47, 117.41 (2s, CN),
113.10 (d, ArC), 86.23, 86.18 (2d, C4’H), 85.95, 85.90 (2s, 2H; ArC),
78.08, 78.04 (2d, C1’), 76.25, 76.11, 75.80, 75.66 (4d, C3’), 64.08, 64.03 (t,
C5’), 58.38, 58.23 (2t, CH2OP), 55.18 (q, OCH3), 43.29, 43.24 (2t, C2’),
43.19, 43.07 (2d, CH-iPr), 24.62, 24.56, 24.49, 24.44 (4q, CH3-iPr), 20.39,
20.34, 20.22, 20.16 (4t, CH2-CN); 31P NMR (202 MHz, CDCl3, 258C): d =
148.57, 148.44; HRMS, LSI-MS: m/z: calcd for C45H52N4O6P1: 775.3625;
found: 775.3612 [M+H]+.
after FC (EtOAc/toluene/TEA 3:6:0.3) as
a white foam. Rf =0.37
(EtOAc/toluene/TEA 3:6:0.2); 1H NMR (300 MHz, CDCl3, 258C): d =
7.62–7.19 (m, 18H; ArH), 6.87 (s, 2H; ArH), 6.84 (s, 2H; ArH), 5.24
(dd, 3J(H,H)=5.5, 9.9 Hz, 1H; C1’H), 4.47 (m, 1H; C3’H), 4.10 (m, 1H;
C4’H), 3.80 (s, 6H; OCH3), 3.40 (dd, 3J(H,H)=4.8, 9.9 Hz, 1H; C5’H),
3.20 (dd, 3J(H,H)=5.5, 9.9 Hz, 1H; C5’H), 2.38 (s, 1H; OH), 2.30 (ddd,
2,3J(H,H)=2.2, 5.9, 7.7 Hz, 1H; C2’H), 2.16–2.09 (m, 1H; C2’H);
1H NMR NOE (400 MHz, CDCl3, 258C): d =5.24 (C1’H) ! 7.4 (ArH;
7.6%), 4.10 (C4’H; 3.5%), 2.30 (C2’H; 5.5%); 4.47 (C3’H) ! 4.10
(C4’H; 2.7%), 3.40 (C5’H; 1.2%), 3.20 (C5’H; 1.4%); 4.10 (C4’H) !
5.24 (C1’H; 4.2%), 4.47 (C3’H; 2.4%), 3.40 (C5’H; 3.0%), 3.20 (C5’H;
1.8%); 2.30 (C2’H)
!
5.24 (C1’H); 10.3%), 4.47 (C3’H; 2.5%);
13C NMR (75 MHz, CDCl3, 258C): d =158.47, 144.86, 140.93, 140.85,
140.50 (5s, ArC), 136.04, 130.13, 129.05, 128.77, 128.24, 128.21, 127.87,
127.25, 127.13, 127.10, 126.82, 126.51, 113.14 (13d, ArC), 86.34 (s, ArC),
86.24 (d, C4’), 79.80 (d, C1’), 74.76 (d, C3’), 64.51 (t, C5’), 55.22 (q,
OCH3), 43.84 (t, C2’); LSIMS: m/z (%): 572 (4) [M+H]+, 303 (100).
Synthesis and purification of the oligonucleotides: All oligonucleotides
were synthesized on a 1 mmol scale on an Applied Biosystems Expedite
Nucleic Acid Synthesizer (8909) using standard phosphoramidite chemis-
try. The phosphoramidites of the natural nucleosides as well as the nu-
cleoside derived CPG solid supports were purchased from Glen Re-
search. The universal support was purchased from CT-Gen (San Jose).
The solvents and reagents used for the synthesis were prepared according
to the manufacturerꢃs indications in the trityl-off mode. The coupling
time for the modified phosphoramidites was extended to 6 min and 2-eth-
ylthio-1H-tetrazole was used as activator. After synthesis, the oligonu-
cleotides were detached and deprotected in concentrated aqueous ammo-
nia (12–18 h at 558C) and filtered through Titan filters (Teflon, 0.45 mm,
Infochroma AB). HPLC was performed on an ꢄkta Basic 10/100 system
(Amersham Pharmacia Biotech). All modified oligonucleotides were
characterized by ESIꢀ-mass spectrometry (Table 6). Concentrations of
oligonucleotides were determined by UV absorption. For the modified
bases, extinction coefficients e260 nm =7100 (bipy) and e260 nm =19000
(biph) were used.
Compound 11b: This compound was prepared as described for 11a, from
1b (30 mg, 0.11 mmol), 4,4’-dimethoxytrityl chloride (49 mg, 0.14 mmol)
in dry pyridine (0.4 mL). Compound 11b (48 mg, 75%) was obtained
after FC (EtOAc/toluene/MeOH/TEA 1:1:0.03:0.03) as a slightly yellow
foam. Rf =0.4 (EtOAc/toluene/MeOH/TEA 1:1:0.03:0.03); 1H NMR
(300 MHz, CDCl3, 258C): d =8.59 (m, 2H; ArH), 8.30 (m, 2H; ArH),
7.82–7.70 (m, 2H; ArH), 7.41–7.39 (m, 2H; ArH), 7.30–7.09 (m, 12H;
ArH), 6.77 (s, 2H; ArH), 6.74 (s, 2H; ArH), 5.19 (dd, 3J(H,H)=5.5,
10.3 Hz, C1’H), 4.39 (m, 1H, C3’H), 4.07 (m, 1H; C4’H), 3.78 (s, 6H;
OCH3), 3.32–3.20 (m, 2H; C5’H), 2.27–2.21 (m, 1H; C2’H) 2.11–2.01 (m,
1H; C2’H); 13C NMR (75 MHz, CDCl3, 258C): d =158.46 (s, ArC),
155.93, 155.39 (2s, ArC), 149.08, 147.20 (2d, ArC), 144.72 (2s, ArC),
137.50 (s, ArC), 136.88 (d, ArC), 135.92 (s, ArC), 134.56 (d, ArC), 130.02,
128.13, 127.80, 126.78 (4d, ArC) 123.60, 121.06, 120.79 (3d, ArC), 113.11
(d, ArC), 86.56 (s, ArC), 86.23 (d, C4’H), 78.97 (d, C1’H), 74.36 (d, C3’),
64.36 (t, C5’), 55.14 (q, CH3), 43.68 (t, C2’); HRMS, LSI-MS: m/z: calcd
for C36H35N2O5: 575.25537; found: 575.25460 [M+H]+.
Compound 12a: N,N-Diisopropylethylamine (213 mL, 1.24 mmol) was
added followed by 2-cyanoethyl diisopropylchlorophosphor amidite
(138 mL, 0.62 mmol) at RT to a solution of 11a (237 mg, 0.41 mmol) in
dry THF (10.6 mL). After 1.5 h CH2Cl2 (40 mL) was added, and the mix-
ture was extracted with saturated aqueous NaHCO3 (3ꢂ20 mL), dried
UV-melting experiments and CD spectra: All UV melting curves were
recorded on a Cary 3E UV/VIS spectrometer (Varian) equipped with a
Peltier block and Varian WinUV software at 260 nm. Unless otherwise
indicated, the oligonucleotide concentration was kept at 1.2 mm in a
buffer solution (10 mm NaH2PO4, 0.15 mm NaCl, pH 7.0) throughout all
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1921

C. J. Leumann et al.
Table 6. MS data and HPLC or FPLC purification of modified oligonucleotides.
Entry
Sequence
Calcd
Found
HPLC Purification
tR
[min]
OD260 nm
[MꢀH]ꢀ
[MꢀH]ꢀ
AE[a]
[%]
RP[c]
[%]
tR
[min]
1
2
3
4
5
6
7
8
5’-CTAGCHHGTCATC-3’
3’-GATCGHHCAGTAG-5’
5’-CTAGCTHGTCATC-3’
3’-GATCGHTCAGTAG-5’
5’-CTAGCGHGTCATC-3’
3’-GATCGHGCAGTAG-5’
5’-CTAGCIHGTCATC-3’
3’-GATCGHICAGTAG-5’
5’-CTAGCYHGTCATC-3’
3’-GATCGHYCAGTAG-5’
5’-CTAGCPHGTCATC-3’
3’-GATCGHPCAGTAG-5’
5’-CTAGHIHITCATC-3’
3’-GATCIHIHAGTAG-5’
5’-CTAGHYHYTCATC-3’
3’-GATCYHYHAGTAG-5’
5’-CTAGHPHPTCATC-3’
3’-GATCPHPHAGTAG-5’
5’-CTAGCIGTCATC-3’
3’-GATCGICAGTAG-5’
5’-CTAGCIIGTCATC-3’
3’-GATCGIICAGTAG-5’
5’-CTAGCIIIGTCATC-3’
3’-GATCGIIICAGTAG-5’
5’-CTAGCIIIIGTCATC-3’
3’-GATCGIIIICAGTAG-5’
5’-CTAGCIIIIIGTCATC-3’
3’-GATCGIIIIICAGTAG-5’
5’-CTAGCIIIIIIGTCATC-3’
3’-GATCGIIIIIICAGTAG-5’
5’-CTAGCIIIIIIIGTCATC-3’
3’-GATCGIIIIIIICAGTAG-5’
5’-TTTTTTTTIIIIIIIIIIIIT-3’
3’-TIIIIIIIIIIIIT-5’
3652.42
3741.48
3776.52
3865.58
3801.53
3890.59
3804.61
3893.67
3806.47
3895.52
3852.65
3941.71
3698.60
3787.66
3702.32
3791.38
3794.68
3883.74
3624.51
3713.57
3956.80
4045.86
4289.09
4378.15
4621.38
4710.44
4953.67
5042.73
5285.96
5375.02
5618.25
5707.31
6663.30
4533.92
4093.26
4142.29
3744.97
3825.02
3626.37
3715.43
3969.52
4049.58
4294.67
4383.74
4628.82
4717.89
4962.98
5052.04
5297.13
5386.19
3958.66
4047.72
3958.66
3652.76
3741.86
3776.25
3865.13
3800.75
3889.88
3804.89
3893.00
3806.74
3895.38
3852.90
3941.65
3698.70
3787.00
3701.63
3790.75
3793.75
3882.88
3624.00
3713.25
3955.88
4045.25
4288.25
4377.25
4621.63
4710.63
4953.25
5042.75
5285.00
5374.13
5617.25
5706.13
6661.75
4532.13
4092.75
4341.5
3744.75
3824.13
3626.25
3715.25
3959.88
4049.00
4294.38
4383.50
4628.38
4718.00
4962.63
5051.63
5296.38
5385.13
3957.88
4046.88
3958.00
40–55 B1*
40–70 B1*
40–55 B1*
40–55 B1*
40–55 B1*
40–60 B1*
20–45 B1[b]
20–45 B1[b]
40–70 B1*
40–70 B1*
20–45 B1[b]
20–45 B1[b]
60–82 B1
60–82 B1
30–65 B1
50–80 B1
40–80 B1
60–95 B1
25–65 B2*
20–60 B2*
60–75 B1
40–95 B1*
50–90 B1*
50–90 B1*
60–95 B1*
60–95 B1*
–
17.3
18.0
17.9
17.5
19.2
19.9
23.0
24.2
23.2
24.6
25.0
25.8
19.5
23.5
26.0
29.3
16.6
24.5
23.2
28.1
16.5
19.2
27.5
27.5
23.0
25.9
–
10–25 B
10–25 B
10–25 B
10–25 B
10–25 B
10–25 B
10–25 B
10–25 B
5–30 B
18.7
19.5
16.4
17.3
16.1
17.0
24.2
25.3
21.0
21.2
24.8
25.4
26.9
27.8
23.4
23.7
27.4
26.7
23.8
24.4
25.6
18.4
27.4
24.8
23.8
23.8
24.0
24.8
28.9
20.0
24.0
24.0
25.0
25.3
23.9
19.2
19.0
23.0
21.0
21.5
16.0
13.7
14.2
14.3
15.3
15.6
24.5
24.6
22.5
23.0
20.1
20.6
20.2
31.4
13.7
30.7
31.7
22.1
30.8
16.1
25.4
16.9
32.0
25.8
24.2
35.7
11.1
20.7
19.6
24.2
29.9
26.0
19.1
38.6
25.9
28.6
18.6
37.1
13.5
30.0
36.3
22.1
18.8
11.2
24.1
29.9
16.5
30.4
36.1
43.0
32.9
n.d.
n.d.
28.8
13.8
12.8
15.2
21.0
10.7
12.0
14.0
20.7
14.1
12.7
10.1
19.8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
5–30 B
10–25 B
10–25 B
10–35 B
10–35 B
10–25 B
10–25 B
10–35 B
10–35 B
10–25 B
10–25 B
10–35 B
10–40 B
10–40 B
10–45 B
20–55 B
20–55 B
20–55 B*
20–55 B*
20–55 B*
30–65 B*
30–65 B*
30–65 B*
30–65 B*
30–65 B*
20–80 B*
40–70 B
20–100 B*
20–80 B*
5–30 B
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
5’-IIIIIIIGTCATC-3’
3’-IIIIIIICAGTAG-5’
5’-CTAIIIIIIATC-3’
3’-GATIIIIIITAG-5’
–
–
–
–
–
–
5’-CTAGCYGTCATC-3’
3’-GATCGYCAGTAG-5’
5’-CTAGCYYGTCATC-3’
3’-GATCGYYCAGTAG-5’
5’-CTAGCYYYGTCATC-3’
3’-GATCGYYYCAGTAG-5’
5’-CTAGCYYYYGTCATC-3’
3’-GATCGYYYYCAGTAG-5’
5’-CTAGCYYYYYGTCATC-3’
3’-GATCGYYYYYCAGTAG-5’
5’-CTAGCYYYYYYGTCATC-3’
3’-GATCGYYYYYYCAGTAG-5’
5’-CTAGCIYGTCATC-3’
3’-GATCGIYCAGTAG-5’
5’-CTAGCYIGTCATC-3’
40–70 B1*
40–70 B1*
50–70 B1
40–60 B1*
30–70 B2*
40–70 B2*
40–95 B2*
40–80 B2*
–
21.5
21.0
16.0
20.7
22.7
22.4
16.2
17.0
–
–
–
–
20.2
22.0
20.6
5–30 B
10–35 B
10–20 B
10–25 B
10–25 B
10–25 B
10–25 B
5–30 B*
5–30 B*
5–40 B*
5–40 B*
10–25 B
10–25 B
10–25 B
–
–
–
40–80 B2*
40–80 B2*
40–80 B2*
[a] AE=Anion-exchange chromatography: Nucleogen DEAE 60–7, 125ꢂ4 mm, with Nucleogen-Guard column 30ꢂ4 mm (Machery–Nagel)for HPLC;
solvent A1 =20 mm KH2PO4, pH 6, in H2O/MeCN 8:2, and solvent B1 =20 mm KH2PO4, 1m KCl in H2O/MeCN 8:2; solvent B2 =20 mm KH2PO4, 1m KCl
in H2O/MeCN 7:3. [b] Mono Q HR 5/5: Pharmacia Biotech for FPLC; solvent A1 and B1. [c] RP=reversed-phase chromatography: Aquapore RP-300,
220ꢂ6 mm (7 mm) with RP-C18 Newguard (7 mm) (Brownlee Labs); solvent A=0.10m (Et3NH)OAc in H2O and solvent B=0.10m (Et3NH)OAc in
H2O/MeCN 1:4. All chromatographic purification were done at a flow rate of 1 mLminꢀ1 and at rt or heated to 608C when indicated with *, and com-
pounds were detected by UV at 260 nm.
measurements. Consecutive heating-cooling-heating cycles in the temper-
ature interval of 0 or 908C or 10 to 908C were applied with a linear gra-
dient of 0.5 8Cminꢀ1. Heating and cooling ramps were in most cases su-
perimposable, exceptions are indicated. Tm values were defined as the maxi-
mum of the first derivative of the melting curve. CD spectra were mea-
sured on a JASCO J-715 spectropolarimeter at the temperature indicated.
1922
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1911 – 1923

Bipyridyl- and Biphenyl-DNA
FULL PAPER
NMR experiments: Temperature dependent 1H NMR spectra of the
duplex 5’d(GATGACIGCTAG)-d(CTAGCIGTCATC) were recorded on
a Bruker DRX 500 spectrometer at 500.13 MHz. The residual H2O was
suppressed by CW presaturation. Proton chemical shifts were referenced
to sodium 3-(trimethylsilyl) 2,2,3,3-[D4]propionate (TSP). NMR data
were processed and analyzed using Bruker software (WinNMR, Version
6.0). The duplex was dissolved in 10 mm NaH2PO4, 150 mm NaCl at pH 7
and repeatedly freeze-dried from D2O prior to spectra acquisition. Final
addition of 100% D2O resulted in a sample-concentration of about 1 mm.
[15] M. Berger, S. D. Luzzi, A.A. Henry, F. E. Romesberg, J. Am. Chem.
Soc. 2002, 124, 1222–1226.
[16] D. L. McMinn, A.K. Ogawa, Y. Wu, J. Liu, P. G. Schultz, F. E. Ro-
mesberg, J. Am. Chem. Soc. 1999, 121, 11585–11586.
[17] A.K. Ogawa, Y. Q. Wu, D. L. McMinn, J. Liu, P. G. Schultz, F. E.
Romesberg, J. Am. Chem. Soc. 2000, 122, 3274–3287.
[18] E. L. Tae, Y. Wu, G. Xia, P. G. Schultz, F. E. Romesberg, J. Am.
Chem. Soc. 2001, 123, 7439–7440.
[19] C. Yu, A.A. Henry, F. E. Romesberg, P. G. Schultz, Angew. Chem.
2002, 114, 3997–4000; Angew. Chem. Int. Ed. 2002, 41, 3841–3844.
[20] S. Atwell, E. Meggers, G. Spraggon, P. G. Schultz, J. Am. Chem. Soc.
2001, 123, 12364–12367.
Gel electrophoresis: 20% Nondenaturing polyacrylamide gels (19:1
mono/bis) were prepared according to standard procedures.[36] DNA sam-
ples (900 pmol) were dried on a Speedvac at low temperature for 2 h,
taken up in the loading buffer (90 mm Tris-borat, 8% saccharose) and
prior to loading on the gel heated up to 908C and then slowly cooled to
48C. The gels were run at 48C and 100 V for 16 h in TBE buffer (90 mm
Tris-borat, pH 7.2). Bands were visualized by UV light.
[21] Y. Wu, A.K. Ogawa, M. Berger, D. L. McMinn, P. G. Schultz, F. E.
Romesberg, J. Am. Chem. Soc. 2000, 122, 7621–7632.
[22] S. Matsuda, A.A. Henry, P. G. Schultz, F. E. Romesberg, J. Am.
Chem. Soc. 2003, 125, 6134–6139.
[23] C. Brotschi, A. Hꢅberli, C. J. Leumann, Angew. Chem. 2001, 113,
3101–3103; Angew. Chem. Int. Ed. 2001, 40, 3012–3014.
[24] C. Brotschi and C. J. Leumann, Angew. Chem. 2003, 115, 1694–
1697; Angew. Chem. Int. Ed. 2003, 42, 1655–1658.
[25] I. Singh, W. Hecker, A.K. Prasad, V. S. Parmar, O. Seitz, Chem.
Commun. 2002, 500–501.
[26] C. Beuck, I. Singh, A. Bhattacharya, W. Hecker, V. S. Parmar, O.
Seitz, E. Weinhold, Angew. Chem. 2003, 115, 4088–4091; Angew.
Chem. Int. Ed. 2003, 42, 3958–3960.
[27] G. A. Kraus, M. T. Molina, J. Org. Chem. 1988, 53, 752–753.
[28] K. Krohn, H. Heins, K. Wielckens, J. Med. Chem. 1992, 35, 511–
517.
[29] W. Timpe, K. Dax, N. Wolf, H. Weidmann, Carbohydr. Res. 1975, 39,
53–60.
[30] A. Klapars, X. H. Huang, S. L. Buchwald, J. Am. Chem. Soc. 2002,
124, 7421–7428.
[31] A. Klapars, S. L. Buchwald, J. Am. Chem. Soc. 2002, 124, 14844–
14845.
[32] P. F. H. Schwab, F. Fleischer, J. Michl, J. Org. Chem. 2002, 67, 443–
449.
[33] C. B. Nielsen, M. Petersen, E. B. Pedersen, P. E. Hansen, U. B.
Christensen, Bioconjugate Chem. 2004, 15, 260–269.
[34] J. Becaud, I. Pompizi, C. J. Leumann, J. Am. Chem. Soc. 2003, 125,
15338–15342.
[1] C. R. Cantor, P. R. Schimmel, in Biophysical Chemistry Part III: The
behavior of biological macromolecules,Vol. 2 (Ed.: P. C. Vapnek),
W. H. Freeman, New York, 1980, pp. 1109–1181.
[2] W. Saenger, Principles of Nucleic Acid Structure, Springer, New
York, 1984.
[3] K. M. Guckian, B. A. Schweitzer, R. X.-F. Ren, C. J. Sheils, P. L.
Paris, D. C. Tahmassebi, E. T. Kool, J. Am. Chem. Soc. 1996, 118,
8182–8183.
[4] K. M. Guckian, B. A. Schweitzer, R. X.-F. Ren, C. J. Sheils, P. L.
Paris, D. C. Tahmassebi, E. T. Kool, J. Am. Chem. Soc. 2000, 112,
2213–2222.
[5] B. A. Schweitzer, E. T. Kool, J. Am. Chem. Soc. 1995, 117, 1863–
1872.
[6] T. J. Matray, E. T. Kool, J. Am. Chem. Soc. 1998, 120, 6191–6192.
[7] S. Moran, R. X.-F. Ren, E. T. Kool, Proc. Natl. Acad. Sci. USA 1997,
94, 10506–10511.
[8] K. M. Guckian, J. C. Morales, E. T. Kool, J. Org. Chem. 1998, 63,
9652–9656.
[9] J. C. Morales, E. T. Kool, Nat. Struct. Biol. 1998, 5, 950–954.
[10] J. C. Delaney, P. T. Henderson, S. A. Helquist, J. C. Morales, J. M.
Essigmann, E. T. Kool, Proc. Natl. Acad. Sci. USA 2003, 100, 4469–
4473.
[11] J. C. Morales, K. M. Guckian, E. T. Kool, Angew. Chem. 2000, 112,
1046–1068; Angew. Chem. Int. Ed. 2000, 39, 990–1009.
[12] T. J. Matray, E. T. Kool, Nature 1999, 399, 704–708.
[13] K. M. Guckian, T. R. Krugh, E. T. Kool, Nat. Struct. Biol. 1998, 5,
954–959.
[35] J. C. Chaput, C. Switzer, Proc. Natl. Acad. Sci. USA 1999, 96,
10614–10619.
[36] J. Sambrook, E. F. Fritsch, T. Maniatis in Molecular Cloning, A Lab-
oratory Manual, Vol. 1, Cold Spring Harbor Laboratory Press, 1989,
pp. 6.36.
[14] M. Berger, A.K. Ogawa, D. L. McMinn, Y. Q. Wu, P. G. Schultz,
F. E. Romesberg, Angew. Chem. 2000, 112, 3069–3071; Angew.
Chem. Int. Ed. 2000, 39, 2940–2942.
Received: August 19, 2004
Published online: January 31, 2005
Chem. Eur. J. 2005, 11, 1911 – 1923
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1923