Synthesis and hybridization properties of oligonucleotides containing polyamines at the C-2 position of purines: A pre-synthetic approach for the incorporation of spermine into oligodeoxynucleotides containing 2-(4,9,13-Triazatridecyl)-2′-deoxyguanosine

Article abstract of DOI:10.1002/1521-3765(20001117)6:22<4188::AID-CHEM4188>3.0.CO;2-N

We have developed a synthesis of spermine-containing oligonucleotides (ODN-sper) which allows incorporation of multiple polyamine residues. This approach was based on the pertrifluoroacetylated 5′DMT-dGsper phosphoramidite synthon. Its coupling yield with

Full text of DOI:10.1002/1521-3765(20001117)6:22<4188::AID-CHEM4188>3.0.CO;2-N

FULL PAPER
Synthesis and Hybridization Properties of Oligonucleotides Containing
Polyamines at the C-2 Position of Purines: A Pre-synthetic Approach
for the Incorporation of Spermine into Oligodeoxynucleotides Containing
2- 4,9,13-Triazatridecyl)-2'-deoxyguanosine
Pierre Potier, Adib Abdennaji, and Jean-Paul Behr*[a]
Abstract: We have developed a synthe-
sis of spermine-containing oligonucleo-
tides ꢀODN-sper) which allows incorpo-
ration of multiple polyamine residues.
This approach was based on the per-
with 1 ± 4 pendent spermines could be
purified and their hybridization proper-
ties were evaluated. Duplex melting
temperatures increased linearly with
the number of polyamine residues
similar initial stability containing other
cation-substituted bases. Moreover, the
stability increase was neither sequence
nor position-dependent, and even con-
tiguous spermine residues did not cross-
talk. Extrapolation based on these find-
ings leads to the conclusion that a duplex
formed with a 22-mer oligonucleotide
containing seven spermine residues
would be as stable as genomic DNA,
which highlights its potential for DNA
strand invasion.
trifluoroacetylated
5'DMT-dGsper
ꢀDT /sperˆ 3.0 Æ 0.28C
in
100mm
m
phosphoramidite synthon. Its coupling
yield with resin-bound ODN decreased
dramatically when close to the 3'-end.
Optimization of the coupling conditions
allowed 22-mer ODNs containing up to
six spermine residues to be synthesized.
Several ODNs of different sequences
NaCl). This compares very favorably
with values reported for duplexes of
Keywords: DNA recognition ´ du-
plex stability
´ minor groove ´
oligonucleotides ´ spermines
Introduction
terms of stabilization of duplex structures, the results were
widely spread and sometimes even opposite, clearly showing
that the introduction of positive charges is not sufficient per
Therapeutic and diagnostic uses of oligonucleotides ꢀODNs)
have attracted the interest of chemists for over 20 years ꢀFor
recent reviews see refs. [1 ± 3]). Numerous modified DNA
structures have been synthesized in order to improve proper-
ties such as affinity for the complementary DNA or RNA
target. An extensive hybridization study including over 200
chemical modifications, however, led to only few compounds
that increased thermal stability.[4] Cationic residues seem an
obvious choice to favor interaction between two polyanions.
This seemed to be the clue for the high melting temperature of
bacteriophage fW-14 DNA[ where half of the thymines are
se. Endogenous and synthetic polyamines are known to
stabilize DNA against thermal denaturation.[
30±32]
This may be
due to formation of a trans-strand network of ammonium
hydrogen bonds involving the O-2 atoms of pyrimidines and
N-3 of purines in the minor groove, thus clipping the DNA
strands together.[
33]
Spermine was therefore attached to an ODN via the C-2
position of purine which bisects the minor groove in duplex
DNA; a strong thermal stabilization of the structure was
5]
[23]
observed. Groove location indeed seems to be of impor-
[6]
[15]
[13]
replaced by a-putrescinylthymidine. Short a-putrescinylth-
tance, since alkylammonium
and imidazolium
groups
ymidine-containing ODNs, however, lost this property.[
7]
A
within the minor groove showed similar effects, whereas
spermine conjugation to the C-4 position of cytosine ꢀwhich
similar strategy, aimed at decreasing repulsion between the
phosphates of the two strands, led to the design of cationic
[8]
[21]
lies in the major groove) actually decreased stability.
[
9]
[10±19]
[34]
[26]
and zwitterionic ODNs, and of ODNs bearing amines,
We
and others
have described the introduction of
polyamines,[
20±27]
guanidines,
[28]
or S-methylthiourea.
[29]
In
spermine into ODNs containing up to two 2-fluoro-2'-
deoxyinosine residues ꢀpost-oligomerization strategy). Un-
fortunately, this method is not suitable for the preparation of
[
a] Prof. Dr. J.-P. Behr, Dr. P. Potier, Dr. A. Abdennaji
Laboratoire de Chimie G e n e tique associ e C.N.R.S.
Universit e Louis Pasteur de Strasbourg
Facult e de Pharmacie, B.P. 24, 67401 Illkirch ꢀFrance)
Fax : ꢀ‡33)388678891
[34]
polysubstituted oligonucleotides, ꢀODN-sper) which may
be of interest for diagnostic and gene therapy applications.
Here we describe the synthesis of a guanine phosphoramidite
9 bearing a protected spermine moiety which is compatible
E-mail: behr@aspirine.u-strasbg.fr
4188
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4188 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 22

4
188±4194
O
O-R
N
O-NPE
N
O-NPE
N
N
N
N
N
N
N
N
N
NH
NH2
a
c
d
N
N
NH2
N
F
N
F
HO
TBS-O
R'-O
R'-O
O
90%
O
75%
O
88%
O
1
OH
O-TBS
2
3
R= H
R= NPE
O-R"
4
R= R'= TBS
OH
5
6
R'= H
b
80%
a
e
76%
4' R= H ; R'= TBS
R'= DMT
O-NPE
N
O-NPE
O-NPE
N
N
N
N
N
N
f
g
N
h
N
N
NH
N
NH
N
NH
DMT-O
DMT-O
DMT-O
7
7%
O
64%
O
77%
O
O
O
O
OH
7
OH
8
O
9
O
NC
P
NH
N
N(iPr)2
N
O
CF3
CF3
H2N
N
H
N
N
N
N
F3C
H
F3C
H
O
CF3
O
CF3
Scheme 1. Synthesis of the phosphoramidite 9. a) TBS-Cl, imidazole, DMF; b) NPE-OH, DEAD, PPh
3
, dioxane; c) PVPHF, tBuONO, toluene, À58C;
Pꢀ-OCH CH CN), tetrazole, CH CN.
d) TEA-ꢀHF) , THF; e) DMT-Cl, pyridine; f) spermine, MeOH, 558C; g)CF CO Et, TMG, MeOH; h) ꢀiPr)
3
3
2
2
2
2
3
with oligonucleotide synthesis ꢀa similar pre-synthetic ap-
proach is being developed for obtaining polyamine ± oligo-
nucleotide libraries[ ). Using this strategy, ODN-sper with up
to six spermine residues were synthesized.
strategy has been described.[23, 26] The pre-synthetic approach
requires synthesis of a protected phosphoramidite which is
compatible with automated DNA synthesis. As a frequent
consequence of decreased solubility and reactivity, it also
needs screening and optimization of the automated DNA
synthesis parameters ꢀcoupling efficiencies <90% for the
non-natural nucleotide would lead to inseparable ODN
mixtures).
35]
Results and Discussion
The post-synthetic strategy is more versatile. It relies on the
introduction of a nucleotide bearing a small reactive group
such as a halogen atom. ODN synthesis thus requires no
optimization and so a single tagged ODN can serve for the
introduction of various molecular groups. 2-Fluoroinosine is
Synthetic strategy: The introduction of molecular groups into
ODNs can be carried out either before ꢀthe pre-synthetic
approach) or after ꢀthe post-synthetic, or convertible nucleo-
tide approach[36]) the stepwise ODN synthesis. Both methods
have been used to achieve conjugation of polyamines to
ODNs. For C-2-spermine-derivatized purines, only the latter
an ideal precursor for the preparation of guanine N-2
adducts.[
37, 38]
ODN-sper with up to two spermine-derivatized
guanines ꢀGsper) have initially been obtained from oligonu-
Abstract in French: Nous avons d e velopp e une m e thode de
synth eÁ se d'oligonucl e otides contenant des r e sidus spermine
cleotides containing this halogenated nucleotide.[
23, 26]
ODNs
with up to seven fluoroinosines were obtained subsequent-
ly,[ but attempts to convert them to spermine conjugates led
to complex mixtures. We were therefore bound to synthesize
the spermine-deoxyguanosine phosphoramidite 9 and to
optimize its incorporation into an oligonucleotide.
ꢀ
ODN-sper)permettant lꢁincorporation de re Âsidus multiples.
34]
Cette approche est bas e e sur le phosphoramidite du synthon
per-trifluoroacetyl 5'DMT-dGsper. Son rendement de couplage
avec un oligonucl e otide ꢀODN)porte  par la r e sine d e croît
dramatiquement au fur et aÁ mesure que l'on se rapproche de
lꢁextr e mit e 3'. L'optimisation des conditions de couplage a
n e anmoins permis de synth e tiser des ODN 22-m eÁ res contenant
jusqu' aÁ six r e sidus spermine. Plusieurs ODN's de s e quences
vari e es, contenant de une aÁ quatre pendigouillantes spermines
ont e t e purifi e s et e tudi e s. Les temp e ratures de fusion des
hybrides bicat e naires correspondants augmentent proportio-
Synthesis of phosphoramidite 9: Fluorine displacement is an
efficient method for introducing nucleophilic residues at the
C-2 position of purine nucleosides.[
39±45]
We followed a similar
route, using the key 2-fluoro-2'-deoxyinosine intermediate 5
Scheme 1). Deoxyribose protection of 1 was achieved using
ꢀ
tert-butyldimethylsilylchloride. The O-6 was protected to
increase subsequent fluorination yields. Compound 2 was
thus treated with p-nitrophenylethyl alcohol using Pfleidererꢂs
procedure[ to yield compound 3. As described earlier,
several methods were tested for diazotation and fluorination
of deoxyguanosine.[ The best results were obtained in
nonaqueous medium using tert-butylnitrite and polyvinylpyr-
idinium/HF[ as source of fluorine. The reaction proceeded
so smoothly that the silyl groups remained in place: the main
product was the disilylated 2-fluoro-2'-deoxyinosine ꢀ4), with
monoprotected derivative 4' as by-product. The latter com-
pound was resilylated to afford the protected fluoro-2'-
nellement au nombre de r e sidus polyamine ꢀDT /sper ˆ 3.0 Æ
m
0
.28C dans 100mm NaCl). Ces valeurs se comparent avanta-
geusement aÁ celles d e crites pour des s e quences de stabilit e s
initiales comparables, comprenant d'autres substitutions catio-
niques sur les bases. De plus, le gain de stabilit e ne d e pend ni de
la s e quence ni de la position, et m eà me des spermines contigu eÈ s
ne se g eà nent pas. Une extrapolation bas e e sur ces r e sultats
conduit aÁ la conclusion qu'un oligonucl e otide 22 m eÁ re
contenant sept r e sidus spermine serait aussi stable que lꢁADN
g e nomique, r e v e lant ainsi le potentiel de cette approche pour
lꢁinvasion de brin dꢁADN.
46]
34]
47]
Chem. Eur. J. 2000, 6, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4189 $ 17.50+.50/0
4189

J.-P. Behr et al.
FULL PAPER
deoxyinosine ꢀ4) with an overall yield of 75%. The lipophilic
TBS groups allowed straightforward purification of 4. The
silyl groups were gently removed with triethylamine trisꢀhy-
drogen fluoride)[ ꢀ5) and the primary alcohol was selectively
protected with DMT chloride, providing the fluorodeoxyino-
sine derivative 6.
Table 1. Optimization of the coupling yield for reaction of 9with a nascent
ODN.
Conditions
I
II
III
IV
V
48]
solvent
CH
2
Cl
2
/CH
3
CN CH
3
CN CH
3
CN CH
3
3
CN CH CN
concentration
number of steps
coupling time
coupling yield
0.1m
1
10 min
61.0%
0.09m
1
0.09m
1
0.09m
1
0.09m
2
10 min 15 min 20 min 10 min
81.8% 82.9% 83.0% 94.8%
Spermine was readily incorporated into the purine nucleo-
side through its primary amines by heating the reaction
mixture in methanol,[ leading to the 2-spermino-2'-deoxy-
inosine ꢀ7). The excess of hydrophilic spermine was removed
49]
CH CN mixtures), coupling times and concentrations. Di-
3
by extraction in a ternary solvent system ꢀCH Cl /iPrOH/
2
2
chloromethane, which allowed higher concentrations to be
used, let to decreased yields possibly because of reduced
acidity of tetrazole. Coupling times larger than 10 min did not
improve yields. The best results were obtained with two
coupling steps, using a 0.09m solution of 9 in acetonitrile. We
also observed variable coupling yields for 9 along the ODN
synthesis, in otherwise identical conditions. A plot of yield
versus coupling step ꢀFigure 2) clearly shows that the closer to
water) and no further purification was necessary. The
trifluoroacetyl group ꢀTFA) seemed to be a good polyamine
protecting group for ODN synthesis[
21, 24, 27]
but pertrifluor-
oacetylation of 7 proved to be tedious. Trifluoroacetic
anhydride could not be used as it causes glycosidic bond
cleavage. Smoother methods using ethyl trifluoroacetate or S-
ethyl trifluorothioacetate in the presence of triethylamine
gave rather poor yields. Replacement of triethylamine by
other bases has been shown to significantly improve the
protection of some amino acids.[ Using tetramethylguani-
dine ꢀTMG) as base, ethyl trifluoroacetate converted 7 into
the tris protected compound 8 with 64% yield. TMG also
reacted with ethyl trifluoroacetate, so the reagents had to be
used in large excess. Prior to chromatography, the TMG ±
TFA adduct had to be removed by washing the crude material
with diethyl ether. Phosphitylation using a standard proto-
50]
[
51]
col
graphic purification, the two phosphorus diastereoisomers 9a
fast migrating) and 9b ꢀslow migrating) were separated
afforded the phosphoramidite 9. During chromato-
ꢀ
31
incidentally ꢀsee P NMR in Figure 1), although mixed for
the coupling reaction.
Figure 2. Coupling yields decrease with decreasing distance resin/com-
pound.
the 3'-end, the worse the yield. The simplest explanation is
restricted approach of the large phosphoramidite 9 to the
growing oligonucleotide chain when close to the CGP-resin.
We still were able to incorporate up to six Gsper with good
yields ꢀTable 2), yet subsequent work-up was unsuccessful up
to now for ODN-sper with more than four Gsper.
Figure 1. 31P NMR spectra of A) isomer 9a, B) mixture of isomers 9a and
9
b.
Table 2. Oligonucleotide sequences. X: 2-spermino-2'-deoxyinosine.
Oligodeoxynucleotide synthesis: ODNs were synthesized on
controlled pore glass using standard phosphoramidite chem-
istry.[ The behavior of synthon 9 in the ODN synthesis
conditions was checked by TLC. The inertness of the
Name
Sequence
52]
2
2
2dGsp0-a
2dGsp1-a
5'-ATG AGA TGT GAC GAA CGT GTA C-3'
5'-ATG AGA TGT GAC GAA CGT XTA C-3'
5'-ATG AXA TGT GAC XAA CXT GTA C-3'
5'-ATG AXA TXT XAC XAA CXT XTA C-3'
5'-GTA CAC GTT CGT CAC ATC TCA T-3'
5'-TGG TAA AAT GGA AGA CGC CAA A-3'
5'-TGX TAA AAT XGA AXA CGC CAA A-3'
5'-TTT GGC GTC TTC CAT TTT ACC A-3'
5'-ATG AAG AGA TAC GCC CTG GTT C-3'
5'-ATX AAX AGA TAC GCC CTX XTT C-3'
5'-GAA CCA GGG CGT ATC TCT TCA T-3'
trifluoroacetyl group was tested, as it was reported that it
22dGsp3-a
22dGsp6-a
2
2
2
22compl-b
22dGsp0-c
2
could undergo transamidation during the capping step.[53]
2compl-a
2dGsp0-b
2dGsp3-b
Mass spectrometry showed no trifluoroacetyl/acetyl exchange
in the capping conditions. The phosphoramidite 9appeared to
be less reactive than the natural base phosphoramidites,
possibly because of steric hindrance due to the large tris-TFA-
spermino moiety. We optimized the coupling reaction ꢀTa-
ble 1) using several solvents ꢀCH CN, CH Cl and CH Cl /
2dGsp4-c
2compl-c
2
3
2
2
2
2
4190
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4190 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 22

Oligonucleotides
4188±4194
Removal of NPE cannot be performed by standard
ammonia ODN deprotection conditions,[
54]
so the solid
support was first treated with DBU in hot pyridine. Thymine
was added as a scavenger, as acrylonitrile which was simulta-
neously liberated from the phosphate cyanoethylester groups
damages DNA.[ ODN-sper were then cleaved and further
deprotected in hot aqueous ammonia. The DMT-on ODN-
sper were purified by reverse-phase HPLC. After DMT
removal, ODN-sper showed a single HPLC peak and were
characterized by MALDI-TOF mass spectrometry. The mass
spectrum of 22dGsp3-a showed a second peak that was
attributed to sequences containing only two Gsper ꢀi.e.,
55]
21dGsp2-a). Indeed, 22dGsp3-a was synthesized with non-
optimal conditions I ꢀTable 1) which led to 21dGsp2-a as a
consequence of incomplete capping reactions.
PAGE electrophoresis of ODN-sper: ODN-sper were 5'-radio-
labeled with 32PꢀATP). The kinase activity was not inhibited
by the pendent polyamine, even for ODN-sper with Gsper
close to the 5'-end ꢀi.e., 22dGsp3-b and 22dGsp4-c). Purity
was confirmed by denaturating PAGE ꢀFigure 3). ODN-sper
had lower electrophoretic mobility than their control ODNs.
The relative retardation was proportional to the number of
spermine substitutions ꢀFigure 3). The mobilities of 22dGsp3-
a and 21dGsp2-a were very different ꢀFigure 3A, lane 3) and
allowed final purification to be performed.
Figure 4. Duplex formation of 22dGsp4-c with the complementary strand
22compl-c.
Table 3. Melting temperatures [8C] of duplexes formed by ODN and by
ODN-sper with their complementary strands in 10mm phosphate buffer
ꢀpH 7.4), 100mm NaCl.
ODN-sper
ODN T
m
ODN-sper T
m
DT
m
m
DT /sper
22dGsp1-a
22dGsp3-a
22dGsp3-b
22dGsp4-c
66.4
66.4
64.0
64.5
69.4
76.1
72.8
76.0
‡ 3.0
‡ 9.7
‡ 3.0
‡ 3.2
‡ 2.9
‡ 2.9
‡ 8.8
‡ 11.5
temperatures increased linearly with the number of spermines
up to four) borne by the oligonucleotide, with an increment
ꢀ
of DT /sperˆ 3.0 Æ 0.28C. This value compares very favor-
m
able with the values reported in similar salt conditions for
ODNs containing other cation-substituted bases. Moreover
ꢀ
Table 3), the stability increase was neither sequence nor
position-dependent with respect to the ODN extremities.
Even contiguous Gsper in 22dGsp4-c did not decrease their
individual contribution to duplex stability. This remarkable
behavior may be a consequence of minor groove location of
the pendent spermine residues, where multiple bidentate
hydrogen-bonding acceptor sites are available ꢀa purine N-3, a
pyrimidine O-2, and two O-4' per base pair) regardless of the
sequence and in both directions. Extrapolation to multiple
spermine substitutions is therefore justified. Stability similar
Figure 3. Electrophoretic mobilities of ODN-sper's. A) Denaturating
PAGE. Lane 1: 22dGsp0-a, 2: 22dGsp1-a, 3: 22dGsp3-a. B) Relative
mobilities decrease with increasing number of spermine modifications.
Duplex formation of ODN-sper: The presence of several
spermine residues linked to the nucleic bases may interfere
with proper duplex formation. Spectroscopic titration of
to that of a large DNA fragment ꢀT ˆ 858C in 100mm NaCl)
m
would be reached with seven spermine residues.
22dGsp4-c ꢀFigure 4) and of control 22dGsp0-c with their
complementary strand 22compl-c was therefore undertaken.
The maximum hypochromicities of 22dGsp0-c and 22dGsp4-c
were similar ꢀ91.1 and 89.6%, respectively) and went through
a minimum for a 1:1 molar ratio; this confirms that ODN-sper
form predominantly duplex structures.
Conclusion
We have succeeded in extending the number of ODN-
conjugated polyamines from two, with the convertible nucleo-
tide approach, to four with the optimized post-synthetic
strategy. The remarkable hybridization properties of ODN-
sper a posteriori justify the choice for purine N-2 conjugation,
which was in turn guided by our initial finding that spermine
was located in the B-DNA minor groove. A reasonable
extrapolation shows that a duplex formed with a 22-mer
Duplex melting temperatures ꢀT ) of ODN-sper with their
m
complementary sequences were measured and compared with
the corresponding natural duplexes ꢀTable 3). A single, clear-
cut transition was observed in all cases. Spermine conjugation
always increased T , although alkyl substitution at N-2
m
interferes with GC base pairing. Unexpectedly, melting
Chem. Eur. J. 2000, 6, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4191 $ 17.50+.50/0
4191

J.-P. Behr et al.
FULL PAPER
oligonucleotide bearing seven spermine residues would be as
stable as genomic DNA. Besides potential solubility problems
ꢀ30 mL, also at À658C) in a Teflon flask. tBuONO ꢀ1.9 mL, 16 mmol) was
gradually added to the mixture under vigorous stirring and the temperature
was slowly allowed to rise to À58C. The reaction was followed by
disappearance of the starting material on TLC. The polymer was filtered
ꢀ
it would be an overall neutral molecule), such an oligonu-
cleotide may be capable of sequence-selective recognition of
double-stranded DNA by strand invasion. The underlying
Watson ± Crick pairing is general ꢀas opposed to triple helix
formation by Hoogsteen pairing with homopurine sequences)
and may find wide applications in diagnosis and therapy.
off and the solution was neutralized with a 0.1m NaHCO
organic layer was dried over MgSO , evaporated, and the residual oil was
purified on a silica gel column eluted with 1 ± 20% MeOH in CH Cl . The
3
solution. The
4
2
2
by-products were re-silylated ꢀTBS-Cl, imidazole, DMF) and purified by
chromatography as described above. All fractions containing the protected
fluoroinosine ꢀ4) were collected, affording the desired product with 75%
1
overall yield. R
f
ꢀMeOH/CH
2
Cl
2
1:99): 0.70; H NMR ꢀ200 MHz, CDCl
3
):
d ˆ 8.25 ꢀs, 1H, H-8), 8.15 ꢀd, 2H, Ar), 7.5 ꢀd, 2H, Ar), 6.4 ꢀt, 1H, H-1'), 4.85
ꢀ
t, 2H, -CH -O), 4.6 ꢀm, 1H, H-3'), 4.0 ꢀm, 1H, H-4'), 3.8 ꢀt, 2H, H-5'), 3.3
2
Experimental Section
ꢀt, 2H, -CH -Ar), 2.6 ꢀm, 1H, H-2'), 2.4 ꢀm, 1H, H-2'), 0.92, 0.88 ꢀ2s, 2 Â
2
1
9
9
H, 2 Â SiC-CH
3
), 0.07, 0.05 ꢀ2s, 2 Â 6H, 2 Â Si-CH
3
);
F NM R
Abbreviations: EtOH: ethanol, TBS: tert-butyldimethylsilyl, DMF: N,N-
dimethylformamide, DBU: 1,8-diaza-bicyclo[5.4.0]undec-7-ene, NPE:
ꢀ188.3 MHz, CD
3
CN, TFA as reference): d ˆ 29.9 ꢀs); FAB-MS ꢀpositive
mode): m/z ꢀ%): calcd for C30
H
46FN
5
O
6
Si
2
ꢀ647.30); found 648.5 ꢀ20)
5): 3',5'-O-Di-
‡
‡
2
-ꢀ4-nitrophenyl)ethyl, DEAD: diethyl azodicaboxylate, PVPHF: poly[4-
vinylpyridinium polyꢀhydrogen fluoride)], tBuONO: tert-butylnitrite,
TEA: triethylamine, TEA-ꢀHF) : triethylamine trisꢀhydrogen fluoride),
DMT: 4,4'-dimethoxytrityl, spermino: -NH-ꢀCH -NH-ꢀCH -NH-
, TFA: trifluoroacetate, TMG: 1,1,3,3-tetramethylguanidine,
Pꢀ-OCH CH CN): 2-cyanoethyl tetraisopropylphosphorodiamidite,
[M‡H] , 304.2 ꢀ100) [M À ribose‡H] .
6-O-[2- 4-Nitrophenyl)ethyl]-2'-deoxy-2-fluoroinosine
3
TBS-6-O-NPE-2'-deoxy-2-fluoroinosine ꢀ4) ꢀ700 mg, 1.08 mmol) was dis-
2
)
3
2
)
4
solved in THF ꢀ6 mL) and mixed with TEA-ꢀHF) ꢀ1.06 mL, 6.48 mmol).
3
ꢀ
ꢀ
CH
iPr)
2
)
3
-NH
2
The mixture was stirred under argon for 17 h at room temperature. The
2
2
2
solvent was evaporated, the residue was dissolved in CH Cl and purified
2
2
TEAA: triethylammoniumacetat, MALDI-TOF: matrix-assisted laser
desorption/ionization ± time-of-flight, PAGE: polyacrylamide gel electro-
phoresis.
2 2
by silica gel column chromatography with 0 ± 5% MeOH in CH Cl . The
solvent was evaporated in vacuo to give a beige foam ꢀ400 mg, 88%). Rf
1
ꢀMeOH/CH Cl 5:95): 0.45;
2
2
H NMR ꢀ300 MHz, CDCl ): d ˆ 8.18 ꢀd, 2H,
3
General: Chemicals and solvents were purchased from Fluka, Aldrich,
Merck, Janssen, Carlo Erba and SDS. TLC analysis was performed on
precoated Merck silica gel 60 F254 plates using phosphomolybdic acid as
indicator. Flash silica gel chromatography was performed with silica Si60
Ar), 8.00 ꢀs, 1H, H-8), 7.50 ꢀd, 2H, Ar), 6.33 ꢀdd, 1H, H-1'), 4.86 ꢀt, 2H,
-CH
2H, -CH
mode): m/z ꢀ%): calcd for C18
[M‡H] , 304.1 ꢀ100) [M À ribose ‡ H]
2
-O), 4.80 ꢀm, 1H, H-3'), 4.20 ꢀm, 1H, H-4'), 3.9 ꢀm, 2H, H-5'), 3.33 ꢀt,
-Ar), 2.96 ꢀm, 1H, H-2'), 2.37 ꢀm, 1H, H-2'); FAB-MS ꢀpositive
ꢀ419.12); found 420.2 ꢀ55)
2
H
5 6
18FN O
1
31
‡
‡
.
ꢀ
40 ± 63 mm, Merck) and the indicated solvent system. H, P NMR spectra
1
9
were recorded on a Bruker DPX-300 spectrometer ꢀ300 MHz) and F on a
Bruker WP-200 SY ꢀ200 MHz).
5'-O- 4,4'-Dimethoxytrityl)-6-O-[2- 4-nitrophenyl)ethyl]-2'-deoxy-2-fluo-
roinosine 6): 6-O-NPE-2'-deoxy-2-fluoroinosine ꢀ5) ꢀ400 mg, 0.95 mmol)
was coevaporated with dry pyridine and the residue was dissolved in
pyridine ꢀ10 mL). DMT chloride ꢀ520 mg, 1.53 mmol) was added and the
solution was kept under argon. After 17 h the reaction was quenched with
MeOH ꢀ10 mL) and the solvent was evaporated. The residue was dissolved
3
',5'-O-Di-tert-butyldimethylsilyl-2'-deoxyguanosine 2): 2'-deoxyguano-
sine ꢀ1) ꢀ1 g, 3.5 mmol), tert-butyldimethylsilyl chloride ꢀ2.1 g, 13.9 mmol),
and imidazole ꢀ1.68 g, 24.7 mmol) were dissolved in DMF ꢀ2 mL) and
stirred at room temperature. After 30 min the solution was clear and a
white solid began to precipitate. After 24 h the reaction was quenched with
ethanol ꢀ10 mL) and stirred for an additional 30 min. The precipitate was
filtered, washed with cold ethanol, crystallized in ethanol at À208C and
in CH
of TEA in CH
2
Cl
2
and purified by chromatography ꢀsilica gel pretreated with 1%
Cl ) with 1 ± 4% MeOH in CH Cl containing 1% TEA to
ꢀTEA/MeOH/CH Cl 1:5:94):
2
2
2
2
yield the title compound ꢀ520 mg, 76%). R
f
2
2
1
dried under vacuum to give 2 as a fine white powder ꢀ1.56 g, 90%). R
f
0.50; H NMR ꢀ300 MHz, CDCl
3
): d ˆ 8.17 ꢀd, 2H, Ar), 8.05 ꢀs, 1H, H-8),
1
ꢀ
EtOH/CH
2
Cl
2
1:9): 0.50; H NMR ꢀ200 MHz, [D
6
]DMSO): d ˆ 8.1 ꢀd, 2H,
),
-O), 4.6 ꢀm, 1H, H-3'), 3.9 ꢀm, 1H, H-4'), 3.8 ꢀt, 2H, H-5'),
-Ar), 2.6 ꢀm, 1H, H-2'), 2.4 ꢀm, 1H, H-2'), 0.92, 0.88 ꢀ2s, 2 Â
); FAB-MS ꢀpositive
7.48 ꢀd, 2H, Ar), 7.38 ꢀd, 2H, Ar), 7.25 ꢀm, Ar), 6.79 ꢀdd, 4H, Ar), 6.38 ꢀt,
1H, H-1'), 4.83 ꢀt, 2H, -CH -O), 4.66 ꢀm, 1H, H-3'), 4.15 ꢀm, 1H, H-4'), 3.77
ꢀs, 6H, -O-CH ), 3.4 ꢀm, 2H, H-5'), 3.31 ꢀt, 2H, -CH -Ar), 2.7 ꢀm, 1H, H-2'),
2.5 ꢀm, 1H, H-2'); F NMR ꢀ188.3 MHz, CD
CN, TFA as reference): d ˆ
5 8
30.3 ꢀs); FAB-MS ꢀpositive mode): m/z ꢀ%): calcd for C39H36FN O
Ar), 8.0 ꢀs, 1H, H-8), 7.6 ꢀd, 2H, Ar), 6.3 ꢀt, 1H, H-1'), 4.85 ꢀbrs, 2H, -NH
2
2
4
3
9
.7 ꢀt, 2H, -CH
.3 ꢀt, 2H, -CH
H, 2 Â SiC-CH
2
3
2
1
9
2
3
3
), 0.07, 0.05 ꢀ2s, 2 Â 6H, 2 Â Si-CH
3
‡
‡
mode): m/z ꢀ%): calcd for C22
H
41
N
5
O
4
Si
2
ꢀ495.27); found 1013.5 ꢀ13)
ꢀ721.25); found 722.2 ꢀ2) [M‡H] , 303.1 ꢀ100) [DMT] .
'-O- 4,4'-Dimethoxytrityl)-6-O-[2- 4-nitrophenyl)ethyl]-2- 3,9,13-triaza-
‡
‡
‡
[
2M‡Na] , 518.3 ꢀ35) [M‡Na] , 496.3 ꢀ11) [M‡H] , 174.1 ꢀ85) [M À
5
‡
‡
ribose‡Na] , 152.1 ꢀ100) [M À ribose ‡ H] .
',5'-O-Di-tert-butyldimethylsilyl-6-O-[2- 4-nitrophenyl)ethyl]-2'-deoxy-
guanosine 3): 3',5'-O-Di-TBS-2'-deoxyguanosine ꢀ2) ꢀ5.8 g, 11.7 mmol),
PPh ꢀ6.15 g, 23.4 mmol), and NPE alcohol ꢀ3.90 g, 23.3 mmol) were
tridecyl)-2'-deoxyguanosine 7): 5'-O-DMT-6-O-NPE-2'-deoxy-2-fluoro-
inosine ꢀ6) ꢀ715 mg, 0.99 mmol) and spermine ꢀ1 g, 4.95 mmol) were
dissolved in MeOH ꢀ40 mL, distilled over magnesium). The reaction
3
2 2
mixture was stirred at 558C under argon for 4 h, then CH Cl ꢀ70 mL) and
3
suspended in dry dioxane ꢀ100 mL). The mixture was stirred for 15 min
under argon and DEAD ꢀ3.7 mL, 23.4 mmol) were added. The suspension
went pink and turned to a clear yellow solution. After 24 h the solvent was
evaporated and the product was purified by silica gel chromatography
iPrOH ꢀ10 mL) were added. The organic layer was washed with water until
its pH was neutral. When separation of the phases was too slow, small
amounts of iPrOH were added. The organic layer was dried over MgSO
4
and evaporated in vacuo. The residue was lyophilized with a benzene/
1
ꢀ
CH
2
Cl
2
/hexane 9:1). The fraction containing the protected compound was
Cl . The desired
ꢀEtOH/CH Cl
MeOH 9:1 mixture providing a white powder ꢀ685 mg, 77%). H NM R
further chromatographed with 1 ± 5% acetone in CH
product was obtained as yellowish foam ꢀ6.01 g, 80%). R
1
2
2
ꢀ300 MHz, CDCl
7.3 ꢀm, Ar), 6.76 ꢀdd, 4H, Ar), 6.26 ꢀt, 1H, H-1'), 4.86 ꢀm, 1H, H-3'), 4.72 ꢀt,
2H, -CH -O), 4.07 ꢀm, 1H, H-4'), 3.79 ꢀs, 6H, -O-CH ), 3.4 ꢀm, 2H, H-5'),
3.27 ꢀt, 2H, -CH -Ar), 2.7 ꢀm, H-2', C-CH -NH-), 1.7 ꢀm, C-CH -C); FAB-
MS ꢀpositive mode): m/z ꢀ%): calcd for C49
3
): d ˆ 8.15 ꢀd, 2H, Ar), 7.71 ꢀs, 1H, H-8), 7.47 ꢀd, 2H, Ar),
f
2
2
1
:9): 0.69; H NMR ꢀ200 MHz, [D
4
]MeOH): d ˆ 8.1 ꢀd, 2H, Ar), 8.0 ꢀs, 1H,
), 4.7 ꢀt, 2H,
-O), 4.6 ꢀm, 1H, H-3'), 3.9 ꢀm, 1H, H-4'), 3.8 ꢀt, 2H, H-5'), 3.3 ꢀt, 2H,
-Ar), 2.6 ꢀm, 1H, H-2'), 2.4 ꢀm, 1H, H-2'), 0.92, 0.88 ꢀ2s, 2 Â 9H, 2 Â
); elemental analysis calcd ꢀ%)
2
3
H-8), 7.6 ꢀd, 2H, Ar), 6.3 ꢀt, 1H, H-1'), 4.85 ꢀbrs, 2H, -NH
2
2
2
2
-
-
CH
CH
2
H
61
N
9
O
8
P ꢀ903.46); found 904.1
‡
‡
‡
ꢀ2) [M‡H] , 602.0 ꢀ5) [M À DMT‡H] , 303.0 ꢀ100) [DMT] .
'-O- 4,4'-Dimethoxytrityl)-6-O-[2- 4-nitrophenyl)ethyl]-2-[tris-N,N',N''-
2
SiC-CH
for C30
1
3
), 0.07, 0.05 ꢀ2s, 2 Â 6H, 2 Â Si-CH
3
5
H
48
N
6
O
6
Si
2
ꢀ644.32): C 55.87, H 7.50, N 13.03; found C 55.71, H 7.31, N
trifluoroacetamido- 3,9,13-triazatridecyl)]-2'-deoxyguanosine 8): 1,1,3,3-
Tetramethylguanidine ꢀ1.4 mL, 11 mmol) and ethyl trifluoroacetate
‡
2.98; FAB-MS ꢀpositive mode): m/z ꢀ%): 1289.6 ꢀ3) [2M‡H] , 645.3 ꢀ22)
‡
‡
[
M‡H] , 301.1 ꢀ100) [M À ribose‡H] .
',5'-O-Di-tert-butyldimethylsilyl-6-O-[2- 4-nitrophenyl)ethyl]-2'-deoxy-2-
fluoroinosine 4): 3',5'-O-Di-TBS-6-O-NPE-2'-deoxyguanosine ꢀ3) ꢀ2.6 g,
.03 mmol) was dissolved in anhydrous toluene ꢀ30 mL) and cooled to
À658C. This solution was added to a suspension of PVPHF ꢀ8 g) in toluene
ꢀ
2.7 mL, 22 mmol) were added to a solution of 5'-O-DMT-6-O-NPE-2-
3
spermino-2'-deoxyinosine ꢀ7) ꢀ680 mg, 0.75 mmol) in MeOH ꢀ25 mL). The
reaction was stirred under argon at room temperature for 17 h, AcOEt
ꢀ12 mL) was added and the solvent was evaporated in vacuo ꢀrepeated
4
twice). The residue was washed with Et
2
O and purified on a neutral
4192
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4192 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 22

Oligonucleotides
4188±4194
aluminum oxide column with 0 ± 15% EtOH in CH
of TEA) to yield the title compound ꢀ575 mg, 64%). R
2
Cl
2
ꢀcontaining 0.2%
ꢀMeOH/CH Cl
2
4
mL aliquots of 125 mm 22compl-c. After each addition the mixture was
f
2
heated 5 min at 908C and slowly cooled to RT. UVabsorption was recorded
at 260 nm with a UVIKON 930 spectrophotometer ꢀKontron Instruments).
Values were corrected and the concentration of single stranded DNA was
normalized after each measurement.
1
5
7
4
ꢀ
:95‡1% TEA): 0.55; H NMR ꢀ300 MHz, CDCl
3
): d ˆ 8.15 ꢀdd, 2H, Ar),
.72 ꢀd, 1H, H-8), 7.47 ꢀd, 2H, Ar), 7.38 ꢀd, 2H, Ar), 7.3 ꢀm, Ar), 6.79 ꢀdd,
H, Ar), 6.33 ꢀt, 1H, H-1'), 4.72 ꢀt, 2H, -CH -O), 4.64 ꢀm, 1H, H-3'), 4.11
m, 1H, H-4'), 3.79 ꢀs, 6H, -O-CH -N-CO- ‡
), 3.4 ꢀm, 16H, H-5' ‡ C-CH
CH -Ar), 2.7 ꢀm, 1H, H-2'), 2.5 ꢀm, 1H, H-2'), 1.9 ꢀm, 4H, C-CH -C), 1.6
2
3
2
Melting temperatures: Hybridization experiments were carried out on a
CARY3 UV-visible spectrophotometer ꢀVarian). Absorption at 260 nm
was recorded versus temperature at a heating rate of 0.58Cmin . Duplex
concentrations were 2mm in a 10mm sodium phosphate buffer, pH 7.4,
containing 100mm NaCl. Transition temperatures were obtained from first-
order derivative plots of absorbance versus temperature. Decane was
added on the top of each sample to avoid evaporation.
-
2
2
ꢀ
C
m, 4H, C-CH
2
-C); FAB-MS ꢀpositive mode): m/z ꢀ%): calcd for
À1
‡
55
H
58
F
9
N
9
‡
O
11
P
ꢀ1191.41); 1192.0 ꢀ4) [M‡H] , 890.0 ꢀ3) [M À
‡
‡
DMT‡H] , 774.0 ꢀ13) [M À ribose‡H] , 303.1 ꢀ100) [DMT] .
5
'-O- 4,4'-Dimethoxytrityl)-6-O-[2- 4-nitrophenyl)ethyl]-2-[tris-N,N',N''-
trifluoroacetamido- 4,9,13-triazatridecyl)]-2'-deoxyguanosine-3'-O- 2-cya-
noethyl-N,N-diisopropyl phosphoramididite 9): Compound 8 ꢀ220 mg,
0
ꢀ
0
.18 mmol) and tetrazole ꢀ12 mg, 0.17 mmol) were dissolved in CH
3 mL) and 2-cyanoethyl-tetraisopropylphosphorodiamidite ꢀ170 mL,
.54 mmol) was added. The solution was stirred under argon at room
2 2
Cl
temperature. After 2 h, the reaction was quenched with EtOH ꢀ2 mL) and
the mixture was immediately evaporated in vacuo. The residue was
Acknowledgement
dissolved in CH
2
Cl
2
ꢀ50 mL) and washed twice with 0.2m NaHCO
and the solvent was
3
This work was supported in part by the Ligue Nationale Contre le Cancer.
We thank Mr. Yves Cordier and Mr. Olivier Rocques for technical support.
We wish also to thank Prof. Dr. Jean-Marie Lehn and Dr. Robert Fuchs for
access to the CARY 3 spectrophotometer and to the PhosphoImager 425.
ꢀ
15 mL). The organic layer was dried over MgSO
4
removed under vacuum. The residue was purified on a silica gel column
with AcOEt/hexane 1:1 to AcOEt. The products were precipitated from
cold hexane and were lyophilized from benzene to give the diastereoisom-
ers 9a and 9b as a beige powder ꢀtotal 209 mg, 77%). Before ODN
synthesis, both isomers were dried under vacuum over P
and then dissolved in the dry solvents indicated below. R
2
O
5
for one week
ꢀAcOEt/hexane
): 9a: d ˆ
CN):
[
1] K. J. Myers, N. M. Dean, Trends Pharmacol. Sci. 2000, 21, 19 ± 23.
f
1
[2] S. Agrawal, R. P. Iyer, Pharmacol. Ther. 1997, 76, 151 ± 160.
[3] A. Alama, F. Barbieri, M. Cagnoli, G. Schettini, Pharmacol. Res. 1997,
3
2
9
:1‡1% of TEA) 9a: 0.48, 9b: 0.32; H NMR ꢀ300 MHz, CDCl
3
3
1
.50, 9b: d ˆ 2.63 ꢀt, 2H, O-C-CH
2
-CN); P NMR ꢀ121.5 MHz, CD
3
3
6, 171 ± 178.
a: d ˆ 149.35, 9b: d ˆ 149.43 ꢀs); FAB-MS ꢀpositive mode): m/z ꢀ%): calcd
‡
[4] S. M. Freier, K. H. Altmann, Nucleic Acids Res. 1997, 25, 4429 ± 4443.
[5] A. M. Kropinski, R. J. Bose, R. A. Warren, Biochemistry 1973, 12,
for C64
H
75
F
9
N
11
O
12P ꢀ1391.52); 1392.5 ꢀ1) [M‡H] , 774.2 ꢀ4) [M À
‡
‡
ribose‡H] , 303.1 ꢀ100) [DMT] .
1
51 ± 157.
Oligonucleotide synthesis: Oligonucleotides were assembled on an auto-
mated DNA synthesizer ꢀApplied Biosystems 380B, CA, USA) using
standard phosphoramidite chemistry. The reagents and phosphoramidites
were purchased from PerSeptive Biosystems and the solid supports from
MilliGen/Biosearch ꢀLCAA-CPG, 1000 Š, 1 mmol). Commercial phos-
phoramidites were dissolved in dry acetonitrile ꢀ0.1m solutions) and 3 min
standard coupling times were used. Several solvents and coupling times
were tested for phosphoramidite 9. Finally, oligonucleotides 22dGsp1-a and
[
[
6] J. S. Wiberg, J. Biol. Chem. 1967, 242, 5824 ± 5829.
7] T. Takeda, K. Ikeda, Y. Mizuno, T. Ueda, Chem. Pharm. Bull. 1987, 35,
3
558 ± 3567.
[
[
8] R. L. Letsinger, C. N. Sigman, G. Histand, M. Salunkhe, J. Am. Chem.
Soc. 1988, 110, 4470 ± 4471.
9] H. Hashimoto, M. G. Nelson, C. Switzer, J. Am. Chem. Soc. 1993, 115,
7
128 ± 7134.
[
[
[
10] J. T. Goodwin, D. G. Lynn, J. Am. Chem. Soc. 1992, 114, 9197 ± 9198.
11] G. Y. Wang, D. E. Bergstrom, Tetrahedron Lett. 1993, 34, 6725 ± 6728.
12] R. Fathi, Q. Huang, J. L. Syi, W. Delaney, A. F. Cook, Bioconjugate
Chem. 1994, 5, 47 ± 57.
13] K. S. Ramasamy, M. Zounes, C. Gonzalez, S. M. Freier, E. A. Lesnik,
L. L. Cummins, R. H. Griffey, B. P. Monia, P. D. Cook, Tetrahedron
Lett. 1994, 35, 215 ± 218.
2
2 2 3
2dGsp3-a were obtained with 0.1m of 9 in 1:1 CH Cl /CH CN but the best
results were obtained using a 0.09m solution in dry acetonitrile. During the
synthesis, the phosphoramidite 9 solution was kept on activated 4 Š
molecular sieves. After introduction of the modified base, the capping step
[
was repeated and, in some cases, the coupling step as well. Resin-supported
À2
ODN-sper were treated with 0.5m DBU in pyridine ꢀcontaining 2 Â 10
m
thymine) for 16 h at 558C, washed twice with dry pyridine, then twice with
water. ODN-sper were cleaved and further deprotected with 32% aqueous
ammonia for 15 min at room temperature and for 16 h at 558C. DMT-
protected oligonucleotides were purified by HPLC on a Uptisphere C18
[
[
[
14] M. Manoharan, K. Tivel, L. K. Andrade, P. Dan Cook, Tetrahedron
Lett. 1995, 36, 3647 ± 3650.
15] M. Manoharan, K. S. Ramasamy, V. Mohan, P. D. Cook, Tetrahedron
Lett. 1996, 37, 7675 ± 7678.
16] R. H. Griffey, B. P. Monia, L. L. Cummins, S. Freier, M. J. Greig, C. J.
Guinosso, E. Lesnik, S. M. Manalili, V. Mohan, S. Owens, B. R. Ross,
H. Sasmor, E. Wancewicz, K. Weiler, P. D. Wheeler, P. Dan Cook, J.
Med. Chem. 1996, 39, 5100 ± 5109.
17] N. Haginoya, A. Ono, Y. Nomura, Y. Ueno, A. Matsuda, Bioconjugate
Chem. 1997, 8, 271 ± 280.
18] Y. Ueno, Y. Nagasawa, I. Sugimoto, N. Kojima, M. Kanazaki, S. Shuto,
A. Matsuda, J. Org. Chem. 1998, 63, 1660 ± 1667.
reverse phase column eluted with increasing concentrations of CH
3
CN in
0
.1m TEAA buffer ꢀpH 7.0). The desired fractions were evaporated and the
residue was treated with 80% aqueous AcOH for 15 min. The solution was
quickly evaporated, 1% aqueous TEA was added and concentrated to give
the fully deprotected oligonucleotides. Samples were further purified by
HPLC ꢀwith the same eluent) to remove minor impurities. MALDI-TOF
MS ꢀnegative-ion mode) m/z: 22dGsp1-a: calcd 7009, found 7008;
[
[
2
2
2dGsp3-a: calcd 7379, found 7376; 21dGsp2-a: calcd 6866, found 6865;
2dGsp3-b: calcd 7381, found 7380; 22dGsp4-c: calcd 7495, found 7488.
[
[
19] N. Mignet, S. M. Gryaznov, Nucleic Acids Res. 1998, 26, 431 ± 438.
20] C. H. Tung, K. J. Breslauer, S. Stein, Nucleic Acids Res. 1993, 21,
PAGE electrophoresis: ODNs were labeled with 32P using T4 polynucleo-
[
56]
À1
tide kinase ꢀBioLabs) and run ꢀ52 Vcm for 2 h at room temperature)
on a denaturating gel ꢀ18.5%/0.5% acrylamide/N,N-methylenebisacryla-
mide, 90mm Tris borate, 2mm EDTA, 7m urea). Labeled ODNs were
visualized with a PhosphoImager 425 ꢀMolecular Dynamics). Relative
5
489 ± 5494.
[
21] T. P. Prakash, D. A. Barawkar, V. Kumar, K. N. Ganesh, Bioorg. Med.
Chem. Lett. 1994, 4, 1733 ± 1738.
22] P. Bigey, G. Pratviel, B. Meunier, Chem. Commun. 1995, 181 ± 182.
23] N. Schmid, J.-P. Behr, Tetrahedron Lett. 1995, 36, 1447 ± 1450.
24] H. Nara, A. Ono, A. Matsuda, Bioconjugate Chem. 1995, 6, 54 ± 61.
25] C. Sund, N. Puri, J. Chattopadhyaya, Tetrahedron 1996, 52, 12275 ±
[
[
[
[
mobilities ꢀm
as reference. 22dGsp1-a: m
ˆ 0.84. Oligonucleotides 21dGsp2-a and 22dGsp3-a were cut out from
the gel, extracted with 0.5m AcONH , 1mm EDTA pH 8.0 buffer and
R
) were measured using the unmodified sequence 22dGsp0-a
R
ˆ 0.91; 21dGsp2-a: m ˆ 0.91; 22dGsp3-a:
R
m
R
4
1
2290.
precipitated with three volumes of cold ethanol.
[
26] A. R. Diaz, R. Eritja, R. G. Garcia, Nucleosides Nucleotides 1997, 16,
Duplex formation experiments: 22dGsp0-c ꢀ2.22 nmol) and 22c-dGsp4
2035 ± 2051.
ꢀ
1.72 nmol) were each dissolved in 10mm phosphate buffer ꢀ1 mL)
[27] Y. Ueno, M. Mikawa, A. Matsuda, Bioconjugate Chem. 1998, 9, 33 ±
containing 100mm NaCl ꢀpH 7.5). Solutions were titrated by adding 1 ±
39.
Chem. Eur. J. 2000, 6, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4193 $ 17.50+.50/0
4193

J.-P. Behr et al.
FULL PAPER
[
28] R. O. Dempcy, K. A. Browne, T. C. Bruice, Proc. Natl. Acad. Sci. USA
995, 92, 6097 ± 6101.
[44] H. Lee, E. Luna, M. Hinz, J. J. Stezowski, A. S. Kiselyov, R. G. Harvey,
J. Org. Chem. 1995, 60, 5604 ± 5613.
1
[
[
[
29] D. P. Arya, T. C. Bruice, J. Am. Chem. Soc. 1998, 120, 6619 ± 6620.
30] H. Tabor, Biochemistry 1962, 1, 496 ± 501.
31] H. S. Basu, H. C. Schwietert, B. G. Feuerstein, L. J. Marton, Biochem.
J. 1990, 269, 329 ± 334.
[45] G. X. He, S. H. Krawczyk, S. Swaminathan, R. G. Shea, J. P. Dough-
erty, T. Terhorst, V. S. Law, L. C. Griffin, S. Coutre, N. Bischofberger,
J. Med. Chem. 1998, 41, 2234 ± 2242.
[46] F. Himmelsbach, B. S. Schulz, T. Trichtinger, R. Charubala, W.
Pfleiderer, Tetrahedron 1984, 40, 59 ± 72.
[
32] D. Esposito, P. Delvecchio, G. Barone, J. Am. Chem. Soc. 1997, 119,
2
606 ± 2613.
[47] G. A. Olah, X.-Y. Li, Synlett 1990, 267 ± 269.
[
[
33] N. Schmid, J. P. Behr, Biochemistry 1991, 30, 4357 ± 4361.
34] A. Abdennaji, P. F. Potier, S. Doronina, I. Huc, J. P. Behr, Tetrahedron
Lett. 1997, 38, 2989 ± 2992.
[48] M. A. McClinton, Aldrichimica Acta 1995, 28, 31 ± 35.
[49] P. Godzina, K. Adrych-Rozek, W. T. Markiewicz, Nucleosides Nu-
cleotides 1999, 18, 2397 ± 2414.
[
35] W. T. Markiewicz, P. Godzina, M. Markiewicz, Nucleosides Nucleo-
tides 1999, 18, 1449 ± 1454.
[50] T. J. Curphey, J. Org. Chem. 1979, 44, 2805 ± 2807.
[51] J. Nielsen, M. Taagaard, J. E. Marugg, J. H. van Boom, O. Dahl,
Nucleic Acids Res. 1986, 14, 7391 ± 7403.
[52] F. Eckstein in Oligonucleotides and analogues: a practical approach
ꢀEds.: D. Rickwood, B. D. Hames), IRL Press, Oxford 1991.
[53] E. Ferrer, G. Neubauer, M. Mann, R. Eritja, J. Chem. Soc. Perkin
Trans. 1 1997, 2051 ± 2057.
[
[
36] A. M. MacMillan, G. L. Verdine, J. Org. Chem. 1990, 55, 5931 ± 5933.
37] C. M. Harris, L. Zhou, E. A. Strand, T. M. Harris, J. Am. Chem. Soc.
1
991, 113, 4328 ± 4329.
38] D. A. Erlanson, L. Chen, G. L. Verdine, J. Am. Chem. Soc. 1993, 115,
2583 ± 12584.
[
1
[
[
39] J. F. Gerster, R. K. Robins, J. Am. Chem. Soc. 1965, 87, 3752.
40] H. Lee, M. Hinz, J. J. Stezowski, R. G. Harvey, Tetrahedron Lett. 1990,
[54] M. Acedo, C. Fabrega, A. Avino, M. Goodman, P. Fagan, D. Wemmer,
R. Eritja, Nucleic Acids Res. 1994, 22, 2982 ± 2989.
3
1, 6773 ± 6776.
41] B. Zajc, M. K. Lakshman, J. M. Sayer, D. M. Jerina, Tetrahedron Lett.
992, 33, 3409 ± 3412.
42] J. Woo, S. T. Sigurdsson, P. B. Hopkins, J. Am. Chem. Soc. 1993, 115,
407 ± 3415.
[55] R. Eritja, J. Robles, A. Avino, F. Albericio, E. Pedroso, Tetrahedron
1992, 48, 4171 ± 4182.
[56] T. Maniatis, E. F. Fritsch, J. Sambrook, Molecular cloning: a laboratory
manual, Cold Spring Harbor Laboratory, New York, 1989.
[
[
[
1
3
43] D. Tsarouhtsis, S. Kuchimanchi, B. L. DeCorte, C. M. Harris, T. M.
Harris, J. Am. Chem. Soc. 1995, 117, 11013 ± 11014.
Received: May 29, 2000 [F2518]
4194
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0622-4194 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 22