In vitro and in vivo activities of oligodeoxynucleotide-based thrombin inhibitors containing neutral formacetal linkages

Article abstract of DOI:10.1021/jm970766i

A series of 15-mer oligodeoxynucleotide analogues were synthesized, and their thrombin inhibitory activities in vitro and in vivo were evaluated. These oligodeoxynucleotide analogues share the same sequence (GGTTGGTGTGGTTGG) but have one or more phosphodi

Full text of DOI:10.1021/jm970766i

4224
J . Med. Chem. 1998, 41, 4224-4231
In Vitr o a n d in Vivo Activities of Oligod eoxyn u cleotid e-Ba sed Th r om bin
In h ibitor s Con ta in in g Neu tr a l F or m a ceta l Lin k a ges
Gong-Xin He,* J ohn P. Williams, Michael J . Postich, S. Swaminathan, Regan G. Shea, Terry Terhorst,
Veronica S. Law, Cheri T. Mao, Cathy Sueoka, Steven Coutre´, and Norbert Bischofberger
Gilead Sciences, 333 Lakeside Drive, Foster City, California 94404
Received August 12, 1998
A series of 15-mer oligodeoxynucleotide analogues were synthesized, and their thrombin
inhibitory activities in vitro and in vivo were evaluated. These oligodeoxynucleotide analogues
share the same sequence (GGTTGGTGTGGTTGG) but have one or more phosphodiester
linkages replaced by a neutral formacetal group. The results obtained from monosubstitutions
show that no single phosphodiester group is critical for the thrombin inhibitory activity,
suggesting that the interaction between the oligodeoxynucleotide and thrombin is based on a
multiple-site charge-charge interaction. Analysis of the effects of different phosphodiester
replacements indicates that the backside and left side of the chairlike structure formed by the
molecule may be involved in binding with thrombin, presumably by having direct contacts
with the anion-binding exosite of the enzyme. For the oligodeoxynucleotides containing two
noncontiguous formacetal groups, the effect of the disubstitution is the sum of the effects
obtained from the corresponding two monosubstitutions. Infusion of an oligodeoxynucleotide
containing four formacetal groups into monkeys showed an increased in vivo anticoagulant
effect and an extended in vivo half-life compared to the unmodified oligodeoxynucleotide.
In tr od u ction
Thrombin is a serine protease with multiple functions
in hemostasis. Its roles in coronary heart disease and
other thrombotic disorders have promoted efforts toward
the identification of specific inhibitors1 as well as the
understanding of their interactions with this enzyme.2
GS-522 is a novel oligodeoxynucleotide, GGTTGGT-
GTGGTTGG (1), discovered through screening of a
combinatorial library containing 1013 96-mer single-
stranded oligodeoxynucleotides (ODNs).3 This 15-mer
ODN is a potent inhibitor of thrombin in vitro and in
vivo with desirable anticoagulant properties.4,5 NMR
studies showed that 1 forms a chairlike structure in
solution consisting of two G-tetrads connected by two
TT loops and a single TGT loop (Figure 1).6-8 Both
NMR7 and structure-activity studies9,10 confirmed the
importance of the tertiary structure for the biological
activity. The anion-binding exosite on the thrombin
molecule has been identified as a putative binding site
for 1.11
Controlled intravenous infusion experiments using
cynomolgus monkeys demonstrated that 1 is a potent
and rapid acting anticoagulant.4,5 The prolongation of
the plasma prothrombin time (PT) is directly propor-
tional to the infusion rate, and the in vivo half-life of 1
is less than 2 min.4 Detailed pharmacokinetic stud-
ies12,13 conducted in cynomolgus monkeys showed that
1 is rapidly cleared from plasma by tissue uptake. The
degradation by nuclease in the blood accounts for only
approximately 22% of the total clearance of 1 in vivo.
F igu r e 1. Schematic drawing of the NMR solution structure
of GGTTGGTGTGGTTGG (1) (top) and the top G-tetrad
(bottom) indicating the conformation of the glycosidic bond.
As a part of the structure-activity relationship study,
we were interested in replacing the charged phosphodi-
ester backbone of 1 with a neutral linkage. One goal of
the study was to derive analogues with a higher
thrombin inhibitory activity, which in practice could
result in a lower dose. Toward that goal, we carried
out the present study to understand which phosphodi-
ester group is directly involved in the interaction with
thrombin. Since 1 binds to thrombin at the anion-
10.1021/jm970766i CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/02/1998

In Vitro, in Vivo Activity of Oligodeoxynucleotides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4225
Ch a r t 1a
Sch em e 1a
a
T, thymine; G, guanine; f, formacetal.
binding exosite which contains a number of positively
charged amino acid residues,11 the negatively charged
phosphodiester groups in the ODN must play an im-
portant role in the binding. Replacing the negatively
charged phosphodiester group with a neutral linkage
will reveal which phosphodiester group or groups are
critical to the interaction. This information will aid in
understanding the interaction of 1 with thrombin and
may help in designing simpler analogues of 1 with
improved properties.
Another goal of this study was to derive analogues
with an increased in vivo half-life through structural
modifications. The short in vivo half-life makes 1 an
ideal anticoagulant for indications such as cardiopul-
monary bypass surgeries, where a quick onset of anti-
coagulation and rapid return to a normal clotting value
are desirable. However, analogues of 1 with an in-
creased in vivo half-life could be useful in clinical
settings, where an extended anticoagulant effect is
necessary (e.g., deep vein thrombosis). Pharmacokinetic
studies showed that the short in vivo half-life of 1 is
mainly due to rapid tissue uptake.12,13 Cellular uptake
of an ODN occurs mainly through receptor-mediated
endocytosis,14,15 and binding of an ODN to the cell
surface is mainly based on charge-charge interac-
tions.16 Therefore, replacing the negatively charged
phosphodiester groups of 1 with a neutral linkage may
decrease the tissue uptake and increase the in vivo half-
life of 1.
In the present study, we selected the formacetal group
as the neutral linkage to replace the charged phos-
phodiester group. The formacetal group is achiral, and
incorporation of the formacetal linkage into ODNs
causes minimal structural perturbations.17,18 In this
paper, we report the synthesis of a series of oligodeoxy-
nucleotide analogues and their thrombin inhibitory
activities in vitro and in vivo. These oligodeoxynucle-
otide analogues have the same sequence as 1 but
contain one or more neutral formacetal groups as the
replacement for phosphodiester linkages (Chart 1).
a
T, thymine; GiBu, N2-isobutyrylguanine; DMT, 4,4′-dimethox-
ytrityl. (i) (Ph)2P(O)OH, NIS, CH2Cl2/Et2O; (ii) 3′-O-benzoylthy-
midine or 3′-O-benzoyl-N2-isobutyryl-2′-deoxyguanosine, TMSOTf,
CH2Cl2, -50 °C; (iii) aq KOH, MeOH-dioxane; (iv) DMT-Cl, Py;
(v) NCC2H4OP(Cl)N(CHMe2)2, EtN(CHMe2)2, CH2Cl2.
CH2-OP(O)R2, in place of a nucleoside phosphate in the
condensation reaction (Scheme 1). The nucleoside phos-
phinate is more stable and can be easily isolated and
purified.20
Neither the phosphate method nor the phosphinate
method works in the synthesis of 2′-deoxyguanosine/2′-
deoxyguanosine dimer (G-f-G). Under van Boom’s
conditions using catalytic amounts of TMSOTf,19 no
reaction occurred, while using excess TMSOTf (2-4
equiv) resulted mainly in silylation of the 5′-OH at low
temperatures (-30 °C) or in extensive decomposition at
higher temperatures (0 °C). By using the method
reported by Matteucci et al.,21,22 which involves the
condensation of a nucleoside methylthiomethyl deriva-
tive, 3′-O-CH2SCH3 (MTM), with the 5′-OH group in the
presence of bromine or N-bromosuccinimide (NBS), a
complicated mixture was obtained. However, we found
that in the presence of 2.5 equiv of trifluoromethane-
sulfonic acid (TfOH), G-f-G can be synthesized in 54%
yield through the condensation using N-iodosuccinimide
(NIS) activation (Scheme 2). The excess acid is used to
minimize the effect of the nucleophilic nitrogens of the
guanine base moiety. Thus, the condensation reaction
of the 3′-O-MTM group with the 5′-OH group is con-
ducted in the presence of 1.5 equiv of NIS and 2.5 equiv
of TfOH under anhydrous conditions at low tempera-
ture, where the purine nucleotide is stable to the acid.
These reaction conditions can also be used to prepare
the protected formacetal-linked dimers containing other
purine nucleosides and pyrimidine nucleosides.23
In this study, dimers T-f-T, T-f-G, and G-f-T were
made by the nucleoside phosphinate method. Dimer
Resu lts
Syn th esis. The formacetal-linked dimers, including
thymidine/thymidine dimer (T-f-T), thymidine/2′-deox-
yguanosine dimer (T-f-G), and 2′-deoxyguanosine/thy-
midine dimer (G-f-T), were prepared through a modified
van Boom’s procedure shown in Scheme 1. In the
original method, van Boom et al.19 used the condensa-
tion of a nucleoside phosphate, 3′-O-CH2-OP(O)(OR)2,
with the 5′-OH group of a second nucleoside in the
presence of trimethylsilyl trifluoromethanesulfonate
(TMSOTf) to synthesize formacetal-linked pyrimidine-
pyrimidine or pyrimidine-purine dinucleotide ana-
logues. We employed a nucleoside phosphinate, 3′-O-

4226 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
He et al.
Sch em e 2a
F igu r e 2. Individual positions and relative PT values of the
oligodeoxynucleotides containing one formacetal linkage.
Ta ble 1. Formacetal Monosubstitutions
no.
oligomer
relative PT
101
102
103
104
105
106
107
108
109
110
111
112
113
114
5′GGT-f-TGGTGTGGTTGG3′
5′GGTTGGTGTGGT-f-TGG3′
5′GGTT-f-GGTGTGGTTGG3′
5′GGTTGGT-f-GTGGTTGG3′
5′GGTTGGTGT-f-GGTTGG3′
5′GGTTGGTGTGGTT-f-GG3′
5′GG-f-TTGGTGTGGTTGG3′
5′GGTTGG-f-TGTGGTTGG3′
5′GGTTGGTG-f-TGGTTGG3′
5′GGTTGGTGTGG-f-TTGG3′
5′G-f-GTTGGTGTGGTTGG3′
5′GGTTG-f-GTGTGGTTGG3′
5′GGTTGGTGTG-f-GTTGG3′
5′GGTTGGTGTGGTTG-f-G3′
1.1 ( 0.1
1.1 ( 0.1
0.7 ( 0.1
0.8 ( 0.1
1.1 ( 0.1
0.6 ( 0.1
0.9 ( 0.1
0.8 ( 0.1
1.1 ( 0.1
0.8 ( 0.1
0.9 ( 0.1
0.8 ( 0.1
0.8 ( 0.1
0.8 ( 0.1
a
GiBu, N2-isobutyrylguanine; DMT, 4,4′-dimethoxytrityl; PAC,
phenoxyacetyl. (i) NIS, 2.5 equiv of TfOH, CH2Cl2/CH3CN, -30
°C; (ii) NH3 in MeOH; (iii) DMT-Cl, Py; (iv) (n-Bu)4NF in THF; (v)
NCC2H4OP(Cl)N(CHMe2)2, EtN(CHMe2)2, CH2Cl2.
days or with different lots of reagents, relative pro-
thrombin time (relative PT) was determined, which is
defined as the activity of the test ODN relative to 1
measured concurrently.
G-f-G was made by the new methylthiomethyl nucleo-
side method. The formacetal-linked dimer units were
incorporated into ODNs with standard solid-phase DNA
chemistry using the phosphoramidite method.24
Th r om bin In h ibitor y Activity. Different from
other thrombin inhibitors, the ODNs described in this
study are not active site binders but inhibit thrombin
by binding to the anionic binding exosite of the thrombin
molecule.11 Binding of the ODN to thrombin does not
inhibit thrombin activity to cleave a small chromogenic
substrate. Therefore, the inhibition constant (Ki) cannot
be determined with the traditional method using a small
chromogenic substrate.
In the present study, a prothrombin time (PT) assay
was used to evaluate the thrombin inhibitory activity
of the ODNs. The assay was conducted by adding
thromboplastin and calcium ion to human plasma to
generate thrombin from prothrombin, and the thrombin
generated catalyzes clotting of fibrinogen.25,26 In the
previous study,9 a thrombin time (TT) assay was used,
where thrombin was added to purified fibrinogen to
cause clotting of fibrinogen. Since the PT assay gener-
ates thrombin from prothrombin, it is more representa-
tive of the in vivo situation. To eliminate the variation
caused by the measurements conducted on different
Figure 2 and Table 1 show the results obtained from
the ODNs where each of the 14 phosphodiester linkages
in 1 was successively replaced with a formacetal linkage.
Replacing the phosphodiester linkage with a neutral
formacetal linkage at T4-f-G5 or T13-f-G14 decreased the
relative PT to 0.7 and 0.6, respectively (103 and 106).
In other positions, however, replacement of the phos-
phodiester linkage by a neutral formacetal linkage had
only a small effect (relative PT ) 0.8) or no effect
(relative PT ) 0.9-1.1).
The results of multiple substitutions are listed in
Table 2. Replacing two phosphodiester linkages simul-
taneously with two formacetal linkages in the molecule
at T3-f-T4 and T12-f-T13 (115) did not affect the relative
PT. However, replacing two phosphodiester linkages
simultaneously at T3-f-T4 and T13-f-G14 (116) decreased
the relative PT to 0.6. The largest decreases were
observed for 117, 119, and 121, where the relative PT
decreased to 0.3-0.4. These three ODNs contain at
least one formacetal linkage in the position of T4-f-G5
or T13-f-G14. These are the two positions where a single
formacetal substitution caused a relatively large effect
on the relative PT as described above. In 122-127,
where two phosphodiester linkages between two G units

In Vitro, in Vivo Activity of Oligodeoxynucleotides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4227
Ta ble 2. Formacetal Polysubstitutions
substitution no.
oligomer
relative PT
di
115 5′GGT-f-TGGTGTGGT-f-TGG3′
116 5′GGT-f-TGGTGTGGTT-f-GG3′
117 5′GGTT-f-GGT-f-GTGGTTGG3′
118 5′GGTT-f-GGTGT-f-GGTTGG3′
119 5′GGTT-f-GGTGTGGTT-f-GG3′
120 5′GGTTGGT-f-GT-f-GGTTGG3′
121 5′GGTTGGTGT-f-GGTT-f-GG3′
122 5′G-f-GTTG-f-GTGTGGTTGG3′
123 5′G-f-GTTGGTGTG-f-GTTGG3′
124 5′G-f-GTTGGTGTGGTTG-f-G3′
125 5′GGTTG-f-GTGTG-f-GTTGG3′
126 5′GGTTG-f-GTGTGGTTG-f-G3′
127 5′GGTTGGTGTG-f-GTTG-f-G3′
128 5′GGT-f-TGGT-f-GTGGT-f-TGG3′
129 5′GGT-f-TGGTGT-f-GGT-f-TGG3′
130 5′GGT-f-TGGT-f-GT-f-GGT-f-TGG3′
131 5′GGT-f-TGG-f-TG-f-TGGT-f-TGG3′
1.1 ( 0.1
0.6 ( 0.1
0.3 ( 0.1
0.7 ( 0.1
0.3 ( 0.1
0.7 ( 0.1
0.4 ( 0.1
0.6 ( 0.1
0.7 ( 0.1
0.7 ( 0.1
0.6 ( 0.1
0.6 ( 0.1
0.8 ( 0.1
0.9 ( 0.1
1.1 ( 0.1
0.5 ( 0.1
0.8 ( 0.1
tri
tetra
F igu r e 3. Time course of the ODN-mediated anticoagulation
(PT) during a constant rate iv infusion of unmodified 1 (b) or
130 (O) in a cynomolgus monkey. The ODNs were infused into
the same animal at an infusion rate of 0.5 mg/kg/min over a
period of 20 min. Blood samples were collected from the animal
at specific time points and assayed.
in the molecule were replaced simultaneously by two
formacetal linkages, the relative PT decreased to 0.6-
0.8. The monosubstitution results showed that replac-
ing one phosphodiester linkage between two G units
decreased the relative PT to 0.8-0.9. It seems that the
effect of the disubstitution is a simple sum of the effects
obtained from the corresponding two individual mono-
substitutions.
the plasma concentration decreased rapidly, and the in
vivo half-life of 130 calculated from the plasma concen-
tration curve was 7.6 min, which is significantly in-
creased from 1.5 min for 1 determined previously by the
same method.12,13
ODNs 128 and 129 contain three formacetal linkages
and have two formacetal linkages simultaneously at T3-
f-T4 and T12-f-T13, respectively. In addition, 128 con-
tains an additional formacetal linkage at T7-f-G8, and
129 contains an additional formacetal linkage at T9-f-
G10. Interestingly, the relative PT values of the two
ODNs were not affected by the trisubstitutions. How-
ever, tetrasubstitution had some effects on the relative
PT value. For example, 130 contains T7-f-G8 and T9-f-
Discu ssion
Mon osu bstitu tion . We have shown that introduc-
ing a base-modified deoxynucleotide residue into 1 can
significantly affect the thrombin inhibitory activity
(relative PT) of the ODN.9,10 Particularly, modifications
at one of the eight G-tetrad-forming 2′-deoxyguanosine
residues (G1, G2, G5, G6, G10, G11, G14, and G15 in Figure
1) produce the largest effect. Modifications stabilizing
the tetrad structure increase the thrombin inhibitory
activity, whereas modifications destabilizing the struc-
ture reduce the activity.9,10 In the present study,
however, replacing any of the 14 phosphodiester link-
ages in 1 with a neutral formacetal linkage does not
dramatically affect the thrombin inhibitory activity.
With the monosubstitutions, the maximum decrease
was observed for T13-f-G14 in 106 (relative PT ) 0.6).
These results are consistent with the previous observa-
tions that the replacement of a phosphodiester linkage
with a formacetal linkage in an ODN causes only
minimal structural perturbations.18 These results fur-
ther suggest that there is not one single phosphodiester
group in 1 which is critical for the binding to thrombin
and that two or more negatively charged phosphodiester
groups may simultaneously be involved. Since the
anion-binding exosite of thrombin is a relatively large
area consisting of multiple positively charged residues,11
it can be reasonably concluded that the binding between
the ODN and thrombin is based on a multiple-site
charge-charge interaction.
G
10 in addition to T3-f-T4 and T12-f-T13 and has a relative
PT value of 0.5.
In Vivo An ticoa gu la tion Stu d y. In the present
study, 130 was chosen for evaluation in an in vivo
anticoagulation study in cynomolgus monkeys. The
ODN can be synthesized through solid-phase synthesis
in the large quantity required by the in vivo study, and
since it contains four formacetal groups, 130 was a good
candidate for studying the effect of the phosphodiester
replacement on the in vivo half-life.
Figure 3 shows the time course of the ODN-mediated
anticoagulation (PT) during a constant rate iv infusion
of 1 and 130, in which four phosphodiester linkages
were replaced simultaneously by four neutral formacetal
linkages at the positions described above. Although 130
had a low relative PT value of 0.5 in vitro, 130 showed
an increased in vivo relative PT of 1.7. Moreover, the
in vivo activity was also extended as shown by the long
tailing in Figure 3. The half-life calculated from the
exponential fitting was about 10 min for 130 compared
to less than 2 min for 1. The plasma concentration of
130 in the blood sample collected during the animal
study was analyzed by HPLC under the conditions
reported previously,13 and a plasma concentration curve
similar to the PT curve was obtained (data not shown),
showing an extended half-life of 130 in plasma. The
plasma concentration of 130 increased rapidly after the
initiation of the infusion and reached a plateau of about
0.1 µM within 2 min under a constant infusion rate of
0.5 mg/kg/min. When the infusion was discontinued,
As shown in Figure 2, all the phosphodiester groups,
where the formacetal substitution causes a decrease in
the activity (relative PT e 0.8), are located on the
backside and left side of the chairlike structure. There-
fore, these two sides may be involved in the binding with
thrombin, possibly, by having direct contacts with the
anion-binding exosite of the enzyme. We found previ-
ously10 that introducing a 1-naphthylmethyl group into

4228 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
He et al.
the N2 position of G6 can increase the thrombin inhibi-
tory activity by 60% in both in vitro and in vivo, and
the result was explained by the increased interaction
with thrombin. The naphthylmethyl group on the N2
position of G6 is located on the backside of the chairlike
structure pointing toward outside (see Figure 1), further
confirming the backside of the chairlike structure as the
probable binding site.
The present results (Figure 2) further indicate that
the two phosphodiester groups between T4 and G5 and
between T13 and G14, where formacetal substitution
causes a relatively large decrease in activity (relative
PT ) 0.6-0.7), are the most important phosphodiester
groups involved in the interaction with thrombin. The
phosphodiester groups between T3 and T4 and between
T12 and T13 may not directly be involved in binding,
since neither monosubstitution nor polysubstitution at
these two positions affects activity.
P olysu bstitu tion . A series of ODNs containing two
noncontiguous formacetal linkages were evaluated. Due
to synthetic difficulties, ODNs containing two or more
contiguous formacetal linkages were not prepared. No
synergistic effect was observed for the disubstitution on
the thrombin inhibitory activity of the ODN. In most
cases, the decrease of the relative PT caused by the
disubstitution is close to the sum of the decreases caused
by the individual monosubstitutions. The lowest rela-
tive PT (0.3-0.4) was observed for 117, 119, and 121,
in which one or two of the phosphodiester groups
between T4 and G5 and between T13 and G14 were
replaced, indicating the importance of these two phos-
phodiester groups.
conclusion that the short in vivo half-life of the ODN is
mainly due to the rapid tissue uptake is further sup-
ported by a pilot study in a cardiopulmonary bypass
(CPB) model in dogs,27 showing that the in vivo half-
life of 1 was significantly increased by bypassing the
lung thereby eliminating a large surface area for the
tissue uptake. As shown in Figure 3, 130 demonstrates
an increased in vivo PT value and an extended thrombin
inhibitory activity compared to 1, even though its in
vitro relative PT is only 0.5. The HPLC analysis of the
plasma concentration of 130 showed an extended half-
life of 130 in plasma, indicating that the increased in
vivo PT value and the extended activity are due to the
extended half-life.
Cellular uptake of ODNs has been extensively stud-
ied.28,29 These studies demonstrate that at low ODN
concentrations (<0.5 µM), cellular uptake of ODNs
occurs mainly through receptor-mediated endocyto-
sis,14,15 involving receptor binding of ODNs on cell
surface as the first step.30 At high concentrations, both
receptor-mediated endocytosis and fluid-phase endocy-
tosis may be involved in the cellular uptake of ODNs.15,28
The receptor binding of ODNs on cell surface is a rapid
and saturatable process and independent of the se-
quence of ODNs.31-33 The binding process is mainly
based on charge-charge interactions between ODN and
receptors.16 As the number of the charges or the length
of the ODN decreases, the binding of ODNs to surface
proteins decreases.34 After endocytosis, most of the
ODN is entrapped in endocytic vesicles, and trafficking
out of the endocytic vesicles into cytoplasm is an
extremely inefficient and slow process.35,36
During the in vivo experiment conducted in this
study, the plasma concentration of the ODN was about
0.1 µM at a constant infusion rate of 0.5 mg/kg/min, and
a receptor-mediated endocytosis process should be the
major pathway for tissue uptake of the ODN. Because
of the nature of the surface binding and subsequent
cellular uptake of ODNs, the slower cellular uptake and
the extended in vivo half-life of 130 are probably due
to the decreased number of the charges of 130 compared
to 1. This hypothesis is consistent with the observation
that the mechanism of cellular uptake was changed
from high-efficiency receptor-mediated endocytosis to
low-efficiency fluid-phase endocytosis by replacing all
negatively charged phosphodiester groups in an ODN
with neutral methylphosphonate groups.37 These data
suggest that it is possible to decrease the cellular uptake
and increase the in vivo half-life of ODNs by replacing
negatively charged phosphodiester groups with neutral
linkages.
Two ODNs (128 and 129) containing three formacetal
linkages and two ODNs (130 and 131) containing four
formacetal linkages were synthesized. All four ODNs
contain both T3-f-T4 and T12-f-T13, the two positions
where replacement of the phosphodiester group was
found to have a minimal effect on activity. Adding
another formacetal substitution, T7-f-G8 in 128 or T9-
f-G10 in 129, to the two disubstitutions caused little
change in the relative PT. However, the tetrasubsti-
tutions in 130 and 131 decrease the activity. These
results can be explained by the accumulation of small
structural perturbations caused by each substitution.
Since the chairlike structure shown in Figure 1 does
not show any change in its UV spectrum upon melting,
we have measured, by using 1H NMR, the melting point
(Tm) of the chairlike structure formed by 130. The
melting point of 130 was about 50 °C compared to 55
°C for 1 determined previously under the same condi-
tions,6 suggesting that introducing multiple formacetal
groups into the ODN results in a destabilization of the
chairlike structure. However, due to the low symmetry
Con clu sion s
1
of the molecule as well as the complexity of the H NMR
A series of 15-mer ODNs with one or more negatively
charged phosphodiester groups replaced by a neutral
formacetal group were synthesized and evaluated for
their thrombin inhibitory activity in vitro and in vivo.
It was found that more than one of the phosphodiester
groups are simultaneously involved in the binding with
thrombin. For the ODNs containing two noncontiguous
formacetal groups, additive results were obtained for
the effect of the substitution on the thrombin inhibitory
activity. The in vivo anticoagulation study in monkeys
showed that the ODN containing four formacetal groups
spectrum, no detailed NMR structural analysis was
conducted.
P h a r m a cok in etics. Detailed in vivo pharmacoki-
netic studies13 conducted in cynomolgus monkeys showed
that 1 is rapidly cleared from plasma by tissue uptake.
The degradation by nucleases in the blood accounts for
only approximately 22% of the total clearance of 1 in
vivo. Replacing the two phosphodiester linkages at the
3′-end of 1 with nuclease-resistant phosphorothioate
linkages has little effect on the in vivo half-life.13 The

In Vitro, in Vivo Activity of Oligodeoxynucleotides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4229
(g90%) and chemical integrity by PAGE, ion-exchange HPLC,
and base composition analysis. Qualitative PAGE analysis
involved inspection of acrylamide gels (25 × 16 × 0.05 cm, 20%
acrylamide, 1:19 cross-linked, 7 M urea, Tris/boric acid/EDTA
buffer) using UV shadowing or staining with Stains-ALL.
Mobilities were compared to oligothymidine standards and
unmodified 15-mer 1. Ion-exchange HPLC analysis was
performed on a Waters 600 E system with UV detection at
254 nm using a Dionex Nucleopac PA-100 column. The elution
gradient was linear from 75 mM LiCl/5 mM LiOH to 1.25 M
LiCl/5 mM LiOH in 25 min. To confirm the presence of the
expected modified nucleoside components, ODNs were com-
pletely digested to nucleoside monomers using snake venom
phosphodiesterase, P1 nuclease, and alkaline phosphatase.
The digested samples were evaluated via reversed-phase
HPLC (Amicon; 10 µm, C18, 25-cm × 4.6-mm column, 37 °C,
2% acetonitrile to 35% acetonitrile in 35 min, constant 50 mM
monobasic potassium phosphate). After correction for extinc-
tion coefficient differences using the factors of 0.608 and 1.00
for G and T, respectively, the peak areas obtained were used
to calculate the relative base composition for each ODN.
ODN extinction coefficients were calculated by the method
reported by Tinoco,39 and unmodified 15-mer 1 has ꢀ ) 143 300
based on the method. The G and T residues present in the
formacetal-linked dimers were treated as normal G and T in
the calculation. Since most of the modified ODNs under
consideration only have one or two phosphodiester linkages
replaced by the formacetal linkages, none of the chromophores
should be significantly affected and the error introduced by
the treatment can be expected to be much less than the
systematic error of the assay.
Syn th esis of F or m a ceta l-Lin k ed Dim er s. Com p ou n d
3a . To a solution of 2a (1 mmol), prepared from thymidine
with published method,40 and diphenylphosphinic acid (0.26
g, 1.2 mmol) in 8 mL of dichloromethane/diethyl ether (2:1,
v/v) was added N-iodosuccinimide (0.27 g, 1.2 mmol) at room
temperature. After 30-min stirring, the mixture was poured
into a mixture of 10% Na2S2O3 (10 mL) and saturated NaHCO3
(10 mL). The resulting mixture was extracted with CH2Cl2
(3 × 10 mL). The organic layer was then washed with brine,
dried over MgSO4, and concentrated. Column chromatography
using SiO2/EtOAc provided 3a as a white foam with 67% yield.
1H NMR (CDCl3): δ 7.4-8.0 (m, 15H, Ar-H), 7.2 (s, 1H, 6-H),
6.24 (dd, J ) 5, 6 Hz, 1H, 1′-H), 5.36 (2dd, J ) 11, 13 Hz, 2H,
-OCH2O-), 4.65 (m, 1H, 3′-H), 4.4 and 4.6 (2dd, J ) 3, 11
Hz, 2H, 5′-Hab), 4.29 (m, 1H, 4′-H), 2.2 and 2.5 (2m, 2H, 2′-
Hab), 1.65 (s, 3H, CH3-).
Com p ou n d 3b. Compound 3b was prepared from com-
pound 2b with the same method as above (yield, 30%). 1H
NMR (CDCl3): δ 12.1 (s, 1H, NH), 11.5 (s, 1H, NH), 7.1-7.8
(m, 16H, Ar-H, 8-H), 6.06 (d, J ) 8 Hz, 1H, 1′-H), 5.58 (dd, J
) 8, 10 Hz, 1H, 3′-H), 5.34 and 5.46 (2dd, J ) 5, 9 Hz, 2H,
-OCH2O-), 4.46 and 4.63 (2dd, J ) 4, 12 Hz, 2H, 5′-Hab),
4.2 (m, 1H, 4′-H), 2.5 and 3.2 (2m, 2H, 2′-Hab), 2.8 (m, 1H,
-CH-), 1.13 (t, J ) 7 Hz, 6H, -CH3).
shows an increased in vivo PT and extended in vivo half-
life compared to the unmodified ODN, suggesting that
replacement of negatively charged phosphodiester groups
with neutral formacetal groups may decrease the tissue
uptake of the ODN.
Exp er im en ta l Section
Gen er a l. Column chromatography was performed using
silica gel 60 from EM Science, and thin-layer chromatography
was performed on Kieselgel 60 F254 aluminum plates. Solvents
used were Burdick & J ackson B&J Chrompure HPLC grade
and dried over 4A molecular sieves. Commercially available
reagents were purchased from Aldrich Chemical unless noted
otherwise. 1H NMR and 2D NMR experiments (COSY,
HMBC, and HMQC) were conducted on a GE QE-300 300-
MHz or Varian Unityplus 500-MHz spectrometer using tet-
ramethylsilane as an internal standard. 31P NMR spectra
were obtained on a General Electric QE-300 instrument using
5% phosphoric acid in D2O as an external standard (in
capillary) for phosphorus. J values are listed in hertz (Hz).
Rela tive P r oth r om bin Tim e (P T). The prothrombin time
(PT) assay was carried out by the standard method4,25 using
citrated plasma (Sigma C7916 or freshly spun at 1500g, 10
min) and thromboplastin/calcium (Sigma T7405) at final
plasma concentrations of 2 and 4 µM for ODNs. In the assay,
the thromboplastin/calcium reagent was added to the plasma
in the presence and absence of an inhibitor, and the time
required for the plasma to form a clot was recorded on a
fibrometer. The assay was performed in duplicate at each
concentration. The clotting time for the control sample was
in the range of 10-13 s. Prothrombin time was the percentage
ratio of the clotting time of test ODN to that of the control
sample. Relative prothrombin time was the activity of test
ODN relative to unmodified 1 measured at the same day and
calculated with the following equation:
relative PT ) (PT1 - 100)/(PT2 - 100)
where PT1 represents the prothrombin time of test ODN and
PT2 represents the prothrombin time of unmodified 1. In the
experiment, PT2 was about 150 at 2 µm and 200 at 4 µm. The
results shown in Tables 1-3 are the average values for the
data measured at different concentrations of ODNs mentioned
above.
Con sta n t Ra te In fu sion of ODN a n d Dose Resp on se
in Vivo. The time course of oligodeoxynucleotide-mediated
anticoagulation during a constant rate iv infusion of unmodi-
fied 15-mer 1 or modified 15-mer 130 was studied using a
cynomolgus monkey by measuring PT. The animal used, as
well as the experiment setting, was the same as those reported
previously.4 The oligodeoxynucleotides were infused into the
same animal at an infusion rate of 0.5 mg/kg/min over a period
of 20 min. Blood samples were collected from the animal at
specific time points and assayed. The plasma concentration
of 130 in the blood samples collected in the experiment was
also analyzed by HPLC using the method reported previ-
ously.13
Syn th esis of Oligom er s. ODN analogues were synthe-
sized by using the standard solid-phase DNA chemistry on
controlled pore glass (CPG) support with the phosphoramidite
method24 on a Biosearch DNA synthesizer. The standard
building blocks suitable for the solid-phase synthesis, 3′-O-
[(diisopropylamino)(cyanoethoxy)phosphino]-5′-O-(4,4′-dimeth-
oxytrityl)-N2-isobutyryl-2′-deoxyguanosine and 3′-O-[(diisopro-
pyla m ino)(cya noet h oxy)phosph in o]-5′-O-(4,4′-dim et hoxy-
trityl)thymidine, as well as CPG supports loaded with the
corresponding building block (G) were purchased from Glen
Research (Sterling, Virginia). The CPG supports loaded with
the modified building block were prepared by the standard
method.38 ODNs were purified by polyacrylamide gel electro-
phoresis (PAGE) (ca. 1 mmol) or reversed-phase HPLC (Sep-
Tech PRP-1 column) (g10 mmol), followed by desalting with
a Sephadex NAP-25 column. ODNs were analyzed for purity
Com p ou n d 4a . Compound 3a (0.8 g, 1.4 mmol) and 3′-O-
benzoylthymidine (0.72 g, 2.1 mmol) were dissolved in 14 mL
of dichloromethane and dried over 4A molecule sieves. After
2-h drying, trimethylsilyl trifluoromethanesulfonate (0.31 g,
1.4 mmol) was added dropwise at room temperature. The
mixture was stirred for 30 min and filtered through Celite into
saturated aqueous NaHCO3. The organic layer was separated,
washed with brine, dried over MgSO4, and concentrated to give
the crude formacetal-linked dimer. The crude mixture was
then redissoved in 20 mL of methanol/dioxane (3:1, v/v), and
the solution obtained was cooled to 0 °C. An aqueous KOH
solution (1 M, 5 mL) was added to the solution. After 30 min,
the reaction solution was neutralized with Dowex acid resin,
filtered, and concentrated under reduced pressure. The
residue was then coevaporated with ethanol (3 × 5 mL)
followed by pyridine (3 × 5 mL). 4,4′-Dimethoxytrityl chloride
(0.45 g, 1.3 mmol) was added to the mixture in 12 mL of
pyridine, and the solution was stirred at room temperature
overnight. The resulting mixture was concentrated under

4230 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
He et al.
reduced pressure. The product was purified with column
chromatography using SiO2/CH2Cl2-MeOH-NEt3 ) 100:5:1
(v/v). Compound 4a was obtained as a white foam (0.56 g,
57%). 1H NMR (CDCl3): δ 7.66 (s, 1H, 6-H), 7.2-7.4 (m, 10H,
6-H + Ar-H), 6.86 (d, J ) 9 Hz, 4H, Ar-H), 6.4 (m, 2H, 1′-
H), 4.79 (dd, AB, J ) 7 Hz, 2H, -OCH2O-), 4.4 (m, 1H, 3′-H),
4.3 (m, 1H, 3′-H), 4.2 (m, 1H, 4′-H), 4.1 (m, 1H, 4′-H), 3.9 (m,
1H, 5′-Ha), 3.81 (s, 3H, -OCH3), 3.7 (m, 1H, 5′-Hb), 3.55 (d, J
) 10 Hz, 1H, 5′-Ha), 3.34 (d, J ) 10 Hz, 1H, 5′-Hb), 2.6 (m,
2H, 2′-Ha), 2.4 (m, 2H, 2′-Ha), 2.3 (m, 2H, 2′-Hb), 2.2 (m, 2H,
2′-Hb), 1.90 (s, 3H, -CH3), 1.51 (s, 3H, -CH3).
Com p ou n d 4b. Compound 4b was prepared from com-
pound 3a and 3′-O-benzoyl-N2-isobutyryl-2′-deoxyguanosine
with the same method as above (yield, 48%). 1H NMR
(CDCl3): δ 7.73 (s, 1H, 8-H), 7.49 (s, 1H, 6-H), 7.1-7.3 (m,
9H, Ar-H), 6.76 (d, J ) 9 Hz, 4H, Ar-H), 6.2 (m, H, 1′-H),
6.1 (m, H, 1′-H), 4.9 (m, 2H, 3′-H), 4.6 (bs, 2H, -OCH2O-),
4.3 (m, 1H, 3′-H), 4.1 (m, 2H, 4′-H), 4.0 (m, 1H, 4′-H), 3.8 (m,
1H, 5′-Ha), 3.71 (s, 6H, -OCH3), 3.6 (m, 1H, 5′-Hb), 3.3 (m,
1H, 5′-Ha), 3.2 (m, 1H, 5′-Hb), 2.8 (m, 1H, -CH-), 2.7 (m,
1H, 2′-Ha), 2.5 (m, 2H, 2′-Hab), 2.2 (m, 1H, 2′-Hb), 1.54 (s,
3H, -CH3), 1.16 (d, J ) 7 Hz, 6H, -CH3).
Com p ou n d 4c. Compound 4c was prepared from com-
pound 3b and 3′-O-benzoylthymidine with the same method
as above (yield, 10%). 1H NMR (CDCl3): δ 7.80 (s, 1H, 8-H),
7.55 (s, 1H, 6-H), 7.2-7.4 (m, 9H, Ar-H), 6.82 (d, J ) 8 Hz,
4H, Ar-H), 6.3 (m, 1H, 1′-H), 6.2 (m, 1H, 1′-H), 4.9 (m, 1H,
3′-H), 4.7 (bs, 2H, -OCH2O-), 4.4 (m, 1H, 3′-H), 4.2 (m, 1H,
4′-H), 4.1 (m, 1H, 4′-H), 3.9 (m, 1H, 5′-Ha), 3.77 (s, 6H,
-OCH3), 3.6 (m, 1H, 5′-Hb), 3.4 (m, 1H, 5′-Ha), 3.3 (m, 1H,
5′-Hb), 2.8 (m, 1H, -CH-), 2.7 (m, 1H, 2′-Ha), 2.5 (m, 2H,
2′-Hab), 2.2 (m, 1H, 2′-Hb), 1.62 (s, 3H, -CH3), 1.22 (d, J ) 7
Hz, 6H, -CH3).
Com p ou n d 5a . 2-Cyanoethyl N,N-diisopropylchlorophos-
phoramidite (0.66 g, 2.2 mmol) was added dropwise to a
solution of compound 4a (1.8 g, 2.2 mmol) and N,N-diisopro-
pylethylamine (0.35 g, 2.7 mmol) in 22 mL of dry CH2Cl2 at 0
°C. The solution obtained was stirred at 0 °C for 30 min and
at room temperature for another 1 h. Methanol, 1 mL, was
added to the solution. After the mixture was stirred at room
temperature for 10 min, 100 mL of CH2Cl2 was added. The
solution obtained was washed with 100 mL of 5% aqueous
NaHCO3 solution and dried over Na2SO4. The solvent was
removed under reduced pressure at room temperature. The
crude product obtained was purified with column chromatog-
raphy using SiO2/EtOAc-MeOH-Et3N ) 100:5:1 (v/v). The
white foam product obtained was redissolved in 10 mL of CH2-
Cl2, and the solution was added dropwise to 100 mL of hexane
under vigorous stirring. The precipitate formed was collected
by filtration and dried under vacuum to give 1.2 g of compound
5a as a white powder with 58% yield. 31P NMR (CDCl3): δ
148.2, 149.0.
using SiO2/CH2Cl2-EtOAc-MeOH ) 50:50:8 (v/v) to give 1.6
g of compound 8 as a pale-yellow foam (yield, 54%). 1H NMR
(DMSO-d6): δ 12.0, 11.7 (2bs, 4H, NH), 8.13, 8.10 (2s, 2H, 8-H),
7.6-6.8 (m, 15H, Ar-H), 6.33, 6.10 (2t, J ) 6.5, 7 Hz, 2H,
1′-H), 4.76 (dd, AB, J ) 17 Hz, 2H, -COCH2OPh), 4.6 (bs, 2H,
-OCH2O-), 4.4 (m, 1H, 3′-H), 4.3 (m, 1H, 3′-H), 4.2 (m, 2H,
5′-Hab), 4.1 (m, 1H, 4′-H), 4.0 (m, 1H, 4′-H), 3.5 (m, 2H, 5′-
Hab), 2.8 (m, 3H, 2′-Ha + -CH-), 2.5 (m, 2′-Hb, 1H), 2.3 (m,
2H, 2′-Hab), 1.1 (m, 12H, -CH3), 1.04 (s, 9H, t-Bu). High-
resolution MS (FAB): calcd for C53H63N10O12Si m/z 1059.4396,
found m/z 1059.4378 (M+ + 1). The formation of 3′-O-CH2-
O-5′ linkage was further confirmed by 2D NMR techniques,
including COSY, HMQC, and HMBC (C-H single-bond and
multibond correlation).
Com p ou n d 9. Compound 8 (1.5 g, 2.8 mmol) was dissolved
in 30 mL of CH2Cl2. Concentrated NH3 in MeOH (60 mL) was
added, and the solution was stirred at room temperature for
30 min. The disappearance of the starting material was
confirmed by TLC using SiO2/CH2Cl2-MeOH ) 100:8 (v/v).
The reaction solution was then concentrated under reduced
pressure and coevaporated with pyridine (3 × 30 mL). The
residue obtained was dissolved in 50 mL of dry pyridine. 4,4′-
Dimethoxytrityl chloride (1.02 g, 3 mmol) was added, and the
resulting solution was stirred at room temperature overnight.
After addition of 5 mL of methanol, the solution was stirred
for an additional 5 min. Then, the reaction solution was
concentrated under reduced pressure. The residue was dis-
solved in 200 mL of CH2Cl2, and the organic solution obtained
was washed with 200 mL of 5% aqueous NaHCO3 and dried
over Na2SO4. Removal of the solvent under reduced pressure
gave the crude product, which was further purified by column
chromatography using SiO2/CH2Cl2-MeOH-Et3N ) 100:5:1
(v/v) to give 1.5 g of the product as a white foam (yield, 87%).
The structure of the product was confirmed by 1H NMR
(DMSO-d6). Next, the product obtained above (1.5 g, 1.2 mmol)
was dissolved in 60 mL of THF. A solution of (n-Bu)4NF in
THF (1 M, 3 mL, 3 mmol) was added to the above solution.
The resulting solution was stirred at room temperature for
30 min. The disappearance of the starting material was
confirmed by TLC using SiO2/CH2Cl2-MeOH-Et3N ) 100:10:1
(v/v). Then, the reaction solution was concentrated under
reduced pressure. The residue was dissolved in 300 mL of
CH2Cl2, and the organic solution obtained was washed with
200 mL of 5% aqueous NaHCO3 and dried over Na2SO4. After
the solvent was removed under reduced pressure, the residue
was purified by column chromatography using SiO2/CH2Cl2-
MeOH-Et3N ) 100:5:1 (v/v) to give 1.0 g of compound 9 as a
white foam (yield, 84%). 1H NMR (DMSO-d6): δ 11.9, 11.6
(2bs, 4H, NH), 8.13, 8.09 (2s, 2H, 8-H), 7.3-7.1 (m, 9H, Ar-
H), 6.8 (d, J ) 7 Hz, 2H, Ar-H), 6.7 (d, J ) 7 Hz, 2H, Ar-H),
6.20, 6.15 (2t, J ) 6.5, 7 Hz, 2H, 1′-H), 5.4 (bs, 1H, OH), 4.76
(dd, AB, J ) 7 Hz, 2H, -OCH2O-), 4.4 (m, 1H, 3′-H), 4.3 (m,
1H, 3′-H), 4.0 (m, 1H, 4′-H), 3.9 (m, 1H, 4′-H), 3.69 (s, 6H,
-OCH3), 3.6 (m, 1H, 5′-Ha), 3.5 (m, 1H, 5′-Hb), 3.2 (m, 2H,
5′-Hab), 2.8 (m, 3H, 2′-Ha + -CH-), 2.5 (m, 2′-Hb, 1H), 2.3
(m, 2H, 2′-Hab), 1.1 (m, 12H, -CH3).
Com p ou n d 5b. Compound 5b was prepared from com-
pound 4b with the same method as above (yield, 33%). 31P
NMR (CDCl3): δ 148.4, 149.1.
Com p ou n d 5c. Compound 5c was prepared from com-
pound 4c with the same method as above (yield, 18%). 31P
NMR (CDCl3): δ 148.9, 149.2.
Com p ou n d 10. Compound 10 was synthesized from com-
pound 9 with the same method as for compound 5a . Com-
pound 10 was obtained as a white powder with 78% yield. 31
NMR (DMSO-d6): δ 148.3, 147.8.
P
Com p ou n d 8. N2-Isobutyryl-3′-O-methylthiomethyl-5′-O-
phenoxyacetyl-2′-deoxyguanosine (1.5 g, 2.8 mmol) and N2-
isobutyryl-3′-O-tert-butyldiphenylsilyl-2′-deoxyguanosine (1.6
g, 2.8 mmol) were dissolved in a solvent mixture containing
20 mL of CH2Cl2 and 20 mL of CH3CN. The solution was dried
over 4A molecular sieves (8-12 mesh beads, 3.0 g) for 1 h and
cooled to -30 °C. TfOH (1.01 g, 6.7 mmol) was added dropwise
followed immediately by NIS (0.72 g, 3.2 mmol). The mixture
was stirred at -30 °C for 1 h. The NIS powder was completely
dissolved, and the solution turned to dark brown. Triethyl-
amine (1.0 g, 10 mmol) was added slowly at -30 °C to
neutralize the acid, and the resulting mixture was then poured
into 200 mL of 5% aqueous Na2S2O3 solution. The mixture
was extracted with CH2Cl2 (100 mL × 2), and the organic
phase was dried over Na2SO4. After the solvent was removed,
the solid residue was purified by column chromatography
Ack n ow led gm en t. This work was supported in part
by SBIR Grant No. 1R43HL484311-01. G.-X.H. ex-
presses gratitude to Dr. J eng-Pyng Shaw, Gilead Sci-
ences Inc., for her valuable comments and suggestions
on in vivo pharmacokinetics of oligodeoxynucleotides.
Refer en ces
(1) Tapparelli, C.; Metternich, R.; Ehrhardt, C.; Cook, N. S. Syn-
thetic Low Molecular Weight Thrombin Inhibitors: Molecular
Design and Pharmacological Profile. Trends Pharmacol. Sci.
1993, 14, 366-376.
(2) Banner, D. W.; Hadavry, P. Crystallographic Analysis at 3.0-Å
Resolution of the Binding to Human Thrombin of Four Active
Site-Directed Inhibitors. J . Biol. Chem. 1991, 266, 20085-20093.

In Vitro, in Vivo Activity of Oligodeoxynucleotides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4231
(3) Bock, L. C.; Griffin, L. C.; Latham, J . A.; Vermaas, E. H.; Toole,
J . J . Selection of Single-Stranded DNA Molecules that Bind and
Inhibit Human Thrombin. Nature 1992, 355, 564-566.
(4) Griffin, L. C.; Tidmarsh, G. F.; Bock, L. C.; Toole, J . J .; Leung,
L. K. In Vivo Anticoagulant Properties of a Novel Nucleotide-
Based Thrombin Inhibitor and Demonstration of Regional
Anticoagulation in Extracorporeal Circuits. Blood 1993, 81,
3271-3276.
(5) Li, W.-X.; Kaplan, A. V.; Grant, G. W.; Toole, J . J .; Leung, L. K.
A Novel Nucleotide-Based Thrombin Inhibitor Inhibits Clot-
Bound Thrombin and Reduces Arterial Platelet Thrombus
Formation. Blood 1994, 83, 677-682.
(6) Wang, K. Y.; McCurdy, S.; Shea, R. G.; Swaminathan, S.; Bolton,
P. H. A DNA Aptamer Which Binds to and Inhibits Thrombin
Exhibits a New Structural Motif for DNA. Biochemistry 1993,
32, 1899-1904.
(7) Wang, K. Y.; Krawczyk, S. H.; Bischofberger, N.; Swaminathan,
S.; Bolton, P. H. The Tertiary Structure of a DNA Aptamer
Which Binds to and Inhibits Thrombin Determines Activity.
Biochemistry 1993, 32, 11285-11292.
(8) Macaya, R. F.; Schultze, P.; Smith, F. W.; Roe, J . A.; Feigon, J .
Thrombin-Binding DNA Aptamer Forms a Unimolecular Qua-
druplex Structure in Solution. Proc. Natl. Acad. Sci. U.S.A. 1993,
90, 3745-3749.
(9) Krawczyk, S. H.; Bischofberger, N.; Griffin, L. C.; Law, V. S.;
Shea, R. G.; Swaminathan, S. Structure-Activity Study of
Oligodeoxynucleotides Which Inhibit Thrombin. Nucleosides
Nucleotides 1995, 14, 1109-1116.
(10) He, G.-X.; Krawczyk, S. H.; Swaminathan, S.; Shea, R. G.;
Dougherty, J . P.; Terhorst, T.; Law, V. S.; Griffin, L. C.; Coutre´,
S.; Bischofberger, N. N2- and C8-Substituted Oligodeoxynucle-
otides with Enhanced Thrombin Inhibitory Activity in vitro and
in vivo. J . Med. Chem. 1998, 41, 2234-2242.
(11) Paborsky, L. R.; McCurdy, S. N.; Griffin, L. C.; Toole, J . J .;
Leung, L. K. The single-stranded DNA Aptamer-binding Site of
Human Thrombin. J . Biol. Chem. 1993, 268, 20808-20811.
(12) Shaw, J .-P.; Fishback, J . A.; Cundy, K. C.; Lee, W. A. A Novel
Oligodeoxynucleotide Inhibitor of Thrombin. I. In vitro Metabolic
Stability in Plasma and Serum. Pharm. Res. 1995, 12, 1937-
1942.
(13) Lee, W. A.; Fishback, J . A.; Shaw, J .-P.; Bock, L. C.; Griffin, L.
C.; Cundy, K. C. A Novel Oligodeoxynucleotide Inhibitor of
Thrombin. II. Pharmacokinetics in the Cynomolgus Monkey.
Pharm. Res. 1995, 12, 1943-1947.
(14) Loke, S. L.; Stein, C. A.; Zhang, X. H.; Mori, K.; Nakanishi, M.;
Subasinghe, C.; Cohen, J . S.; Neckers, L. M. Characterization
of Oligonucleotide Transport into Living Cells. Proc. Natl. Acad.
Sci. U.S.A. 1989, 86, 3474-3478.
(15) Yakubov, L. A.; Deeva, E. A.; Zarytova, V. F.; Ivanova, E. M.;
Ryte, A. S.; Yurchenko, L. V.; Vlassov, V. V. Mechanism of
Oligonucleotide Uptake by Cells: Involvement of Specific Recep-
tors? Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 6454-6458.
(16) Beck, G. F.; Irwin, W. J .; Nicklin, P. L.; Akhtar, S. Interactions
of Phosphodiester and Phosphorothioate Oligonucleotides with
Intestinal Epithelial Caco-2 Cells. Pharm. Res. 1996, 13, 1028-
1036.
(17) Milligan, J . F.; Matteucci, M. D.; Martin, J . C. Current Concepts
in Antisense Drug Design. J . Med. Chem. 1993, 36, 1923-1937.
(18) Gao, X.; Brown, F. K.; J effs, P.; Bischofberger, N.; Lin, K.-Y.;
Pipe, A. J .; Noble, S. A. Probing Structural Factors Stabilizing
Antisense Oligonucleotide Duplexes: NMR Studies of a DNA-
DNA Duplex Containing a Formacetal Linkage. Biochemistry
1992, 31, 6228-6236.
(19) Quaedflieg, P. J . L. M.; van der Marel, G. A.; Kuyl-Yeheskiely,
E.; van Boom, J . H. An Alternative Route to the Preparation of
(3′ f 5′) Methylene Acetal Linked Di- and Trinucleosides. Recl.
Trav. Chim. Pays-Bas 1991, 110, 435-436.
(20) Zhang, J .; Matteucci, M. D. A Mild Method for the Syntheses of
3′-Thioformacetal Dinucleotides: the Development of Diphe-
nylphosphinate as a Glycosyl Donor. Tetrahedron Lett. 1995, 36,
8375-8378.
(21) Matteucci, M. D. Deoxyoligonucleotide Analogues Based on
Formacetal Linkages. Tetrahedron Lett. 1990, 31, 2385-2388.
(22) J ones, R. J .; Lin, K.-Y.; Milligan, J . F.; Wadwani, S.; Matteucci,
M. D. Synthesis and Binding Properties of Pyrimidine Oligode-
oxynucleotide Analogues Containing Neutral Phosphodiester
Replacement: the Formacetal and 3′-Thioformacetal Inter-
nucleoside Linkages. J . Org. Chem. 1993, 58, 2983-2991.
(23) He, G.-X.; Bischofberger, N. Preparation of Formacetal-Linked
Purine-Purine Dinucleotide Analogues. Tetrahedron Lett. 1997,
38, 945-948.
(24) Sinha, N. D.; Biernat, J .; McManus, J .; Koster, H. Polymer
Support Oligonucleotide Synthesis XVIII: Use of â-Cyanoethyl-
N,N-Dialkylamino-/N-Morpholino Phosphoramidite of Deoxy-
nucleosides for the Synthesis of DNA Fragments Simplifying
Deprotection and Isolation of the Final Product. Nucleic Acids
Res. 1984, 11, 4539-4557.
(25) Simmons, A. Hematology, A Combined Theoretical and Technical
Approach, 2nd ed.; Butterworth-Heinemann: Newton, MA, 1997;
Chapter 20, pp 343-360.
(26) White, G. C., II; Marder, V. J .; Colman, R. W.; Hirsh, J .;
Salzman, E. W. In Hemostasis and Thrombosis, Basic Principles
and Clinical Practice, 2nd ed.; Colman, R. W., Hirsh, J ., Marder,
V. J ., Salzman, E. W. J . B., Eds.; Lippincott Company: Phila-
delphia, 1982; Chapter 65, pp 1048-1060.
(27) DeAnda, A., J r.; Coutre, S. E.; Moon, M. R.; Vial, C. M.; Griffin,
L. C.; Law, V. S.; Komeda, M.; Leung, L. L. K.; Miller, D. C.
Pilot Study of the Efficacy of a Thrombin Inhibitor for Use during
Cardiopulmonary Bypass. Ann. Thoracic Surg. 1994, 58, 344-
350.
(28) Vlassov, V. V.; Yakubov, L. A. Oligonucleotides in Cells and in
Organisms: Pharmacological Considerations. In Prospects for
Antisense Nucleic Acid Therapy of Cancer and AIDS; Wickstrom,
E., Ed.; Wiley-Liss, Inc.: New York, 1991; pp 243-266.
(29) Bischofberger, N.; Wagner, R. W. Antisense Approaches to
Antiviral Agents. Semin. Virol. 1992, 3, 57-66.
(30) Shoji, Y.; Doken, K.; Shimada, J .; Mizushima, Y. Characteristic
of Cellur Uptake Mechanism and Subcellular Distribution of
Phosphorothioate Oligodeoxynucleotide in CHRC5 Cells. Drug
Delivery System 1994, 9, 19-24.
(31) Stein, C. A.; Tonkinson, J . L.; Zhang, L.-M.; Yakubov, L.;
Gervasoni, J .; Taub, R.; Rotenberg, S. A. Dynamics of the
Internalization of Phosphodiester Oligodeoxynucleotides in HL60
Cells. Biochemistry 1993, 32, 4855-4861.
(32) Zhao, Q.; Waldschmidt, T.; Fisher, E.; Herrera, C. J .; Krieg, A.
M. Blood 1994, 84, 3660-3666.
(33) Nakai, D.; Seita, T.; Terasaki, T.; Iwasa, S.; Shoji, Y.; Mizushima,
Y.; Sugiyama, Y. Cellular Uptake Mechanism for Oligonucle-
otides: Involvement of Endocytosis in the Uptake of Phosphodi-
ester Oligonucleotides by a Human Colorectal Adenocarcinoma
Cell Line, HCT-15. J . Pharmacol. Exp. Ther. 1996, 278, 1362-
1372.
(34) Yakubov, L.; Khaled, Z.; Zhange, L.-M.; Truneh, A.; Vlassov, V.;
Stein, C. A. Oligodeoxynucleotides Interact with Recombinat
CD4 at Multiple Sites. J . Biol. Chem. 1993, 268, 18818-18823.
(35) Akhtar, S.; Basu, S.; Wickstrom, E.; J uliano, R. L. Interactions
of Antisense DNA Oligonucleotide Analogues with Phospholipid
Membranes (liposomes). Nucleic Acids Res. 1992, 19, 5551-5559.
(36) Tonkinson, J . L.; Stein, C. A. Patterns of Intracellular Com-
partmentalization, Trafficking and Acidification of 5′-Fluorescein
Labeled Phosphodiester and Phosphorothioate Oligodeoxynucle-
otides in HL60 Cells. Nucleic Acids Res. 1994, 22, 4268-4275.
(37) Shoji, Y.; Akhtar, S.; Periasamy, A.; Herman, B.; J uliano, R. L.
Mechanism of Cellular Uptake of Modified Oligodeoxynucle-
otides Containing Methylphosphonate Linkages. Nucleic Acids
Res. 1992, 19, 5543-5550.
(38) Damha, M. J .; Giannaris, P. A.; Zabarylo, S. V. An Improved
Procedure for Derivatization of Controlled-Pore Glass Beads for
Solid-Phase Oligonucleotide Synthesis. Nucleic Acids Res. 1990,
18, 3812-3813.
(39) Borer, P. N. CRC. In The Handbook of Biochemistry and
Molecular Biology, 3rd ed.; Fasman, G. D., Ed.; 1975; Vol. 1, p
589.
(40) Medina, J . C.; Salomon, M.; Kyler, K. S. A mild method for the
conversion of alcohol to methylthiomethyl ethers. Tetrahedron
Lett. 1988, 29, 3773-3776.
J M970766I