Synthesis, structural studies and biological properties of new TBA analogues containing an acyclic nucleotide

Article abstract of DOI:10.1016/j.bmc.2008.07.040

A new modified acyclic nucleoside, namely N¹-(3-hydroxy-2-hydroxymethyl-2-methylpropyl)-thymidine, was synthesized and transformed into a building block useful for oligonucleotide (ON) automated synthesis. A series of modified thrombin binding

Full text of DOI:10.1016/j.bmc.2008.07.040

Bioorganic & Medicinal Chemistry 16 (2008) 8244–8253
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis, structural studies and biological properties of new TBA
analogues containing an acyclic nucleotide
a
a
a
a
Teresa Coppola a, Michela Varra a, , Giorgia Oliviero , Aldo Galeone , Giuliana D’Isa , Luciano Mayol ,
*
Elena Morelli b, Maria-Rosaria Bucci c, Valentina Vellecco c, Giuseppe Cirino c, Nicola Borbone a
a Dipartimento di Chimica delle Sostanze Naturali, Università degli Studi di Napoli ‘‘Federico II”, via D. Montesano 49, 80131 Napoli, Italy
b Dipartimento di Chimica Farmaceutica e Tossicologica, Università degli Studi di Napoli ‘‘Federico II”, via D. Montesano 49, 80131 Napoli, Italy
c Dipartimento di Farmacologia Sperimentale, Università degli Studi di Napoli ‘‘Federico II”, via D. Montesano 49, 80131 Napoli, Italy
a r t i c l e i n f o
a b s t r a c t
A new modified acyclic nucleoside, namely N1-(3-hydroxy-2-hydroxymethyl-2-methylpropyl)-thymi-
dine, was synthesized and transformed into a building block useful for oligonucleotide (ON) automated
synthesis. A series of modified thrombin binding aptamers (TBAs) in which the new acyclic nucleoside
replaces, one at the time, the thymidine residues were then synthesized and characterized by UV, CD,
MS, and 1H NMR. The biological activity of the resulting TBAs was tested by Prothrombin Time assay
(PT assay) and by purified fibrinogen clotting assay. From a structural point of view, nearly all the new
TBA analogues show a similar behavior as the unmodified counterpart, being able to fold into a bimolec-
ular or monomolecular quadruplex structure depending on the nature of monovalent cations (sodium or
potassium) coordinated in the quadruplex core. From the comparison of structural and biological data,
some important structure–activity relationships emerged, particularly when the modification involved
the TT loops. In agreement with previous studies we found that the folding ability of TBA analogues is
more affected by modifications involving positions 4 and 13, rather than positions 3 and 12. On the other
hand, the highest anti-thrombin activities were detected for aptamers containing the modification at T13
or T12 positions, thus indicating that the effects produced by the introduction of the acyclic nucleoside on
the biological activity are not tightly connected with structure stabilities. It is noteworthy that the mod-
ification at T7 produces an ON being more stable and active than the natural TBA.
Article history:
Received 14 February 2008
Revised 9 July 2008
Accepted 16 July 2008
Available online 20 July 2008
Keywords:
TBA analogues
Thrombin inhibitor
Acyclic nucleoside phosphoramidite
CD quadruplex structures
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
human thrombin (EC50 of 20 nM) is the TBA,6–15 a 15-mer oligonu-
cleotide identified in 1992 by SELEX technology16,17 (Systematic
Thrombin is a serine protease playing a central role in the clot-
ting process; converting soluble fibrinogen to fibrin; activating the
factors V, VIII, and XI; and stimulating platelet aggregation. A high
number of bio-molecules bind this protein allowing the fine regu-
lation of the clotting process thus preventing any loss of blood and
avoiding inappropriate clots.1 Alterations between procoagulant
and anticoagulant stimuli are the bases of pathological diseases
such as the formation of venous or arterial thrombi, the major
causes of mortality in the western countries.2–4 For these reasons,
the development of anticoagulant strategies to inhibit thrombo-
genesis is currently of great interest and, among these, the most
important are focused on direct or indirect inhibition of thrombin.
In the latest years, direct thrombin inhibitors have undergone
extensive studies in view of their potential anticoagulant proper-
ties since, differently from heparin, they do not bind to plasma pro-
teins.5 One of these direct inhibitors that specifically binds to
Evolution of Ligands by EXponential enrichment). In vitro studies
showed that TBA, used at nanomolar concentrations, prolongs clot-
ting time from 25 to 43 s in human plasma.6 Notwithstanding its
in vivo half-life is 1–2 min, studies conducted on canine cardiopul-
monary bypass showed that it is able to maintain extracorporeal
circulation by employing a constant infusion.18
The solution-state three-dimensional structure of the TBA has
been determined by NMR methods.12–14,19,20 Feigon and co-work-
ers12 have evidenced that their NMR data fitted with both mono-
molecular and symmetrical intermolecular dimeric structure. On
the basis of other NMR results produced by studying the folding
of modified aptamers, they deduced that the tertiary structure of
the TBA was intramolecular, so that, it is generally assumed that
it adopts a chair-like structure consisting of two G-tetrads con-
nected by two lateral TT loops and a central TGT loop (a, Fig. 1).
After this work the intermolecular folding (b, Fig. 1) was not
further investigated. However, Vorlicôva21 and co-workers have
recently demonstrated by CD analyses and EMSA experiments that
the folding of TBA is bimolecular, putting new questions about
* Corresponding author. Tel.: +39 081678540; fax: +39 081678746.
E-mail address: varra@unina.it (M. Varra).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.07.040

T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
8245
Finally, their thrombin inhibitor activity has been evaluated by PT
and purified fibrinogen clotting assays.
2. Results
2.1. Chemistry
The synthesis of acyclic T derivative phosphoramidite is re-
ported in Scheme 1. The key reaction of the global synthesis is
the Mitsunobu condensation30,31 between 5-hydroxymethyl-
2,2,5-trimethyl-1,3-dioxane (1) and N3-benzoylthymine (2). To
perform this reaction the linker and the base were first submitted
to the pertinent protecting procedure.
The commercially available starting material 1,1,1-
tris(hydroxymethyl)ethane was converted into 5-hydroxymethyl-
2,2,5-trimethyl-1,3-dioxane (1) by reaction with 2,2-dimethoxy-
propane in the presence of pyridinium-p-toluene sulfonate as
catalyst.
Figure 1. Schematic representation of TBA structures: (a) monomolecular quadru-
plex; (b) symmetrical bimolecular quadruplex. In both schematic representations
red and blue squares indicate the G bases in anti and syn conformation, respectively.
The green squares indicate the T4 and T13 bases bound to each other by a hydrogen
bond and stacked on the adjacent G-quartet.
The selective protection of thymine base at N3 position has been
firstly obtained by the Reese32 procedure. The latter consists in the
conversion of thymine into N1,N3-dibenzoyl derivative by using
benzoylchoride (2.2 equiv) and a mixture of acetonitrile-pyridine
(5:2 v/v) as solvent; the resulting product was, in turn, hydrolyzed
by a rapid treatment with a carbonate solution, to obtain the de-
sired N3-benzoylthymine. However this procedure gives rather
poor yields with the recovery of a high quantity of unreacted start-
ing material. In an effort to improve the yields, we found that by
extending the time of the first reaction an increasing amount of
N3-benzoylthymine is achieved probably derived from the decom-
position of the N1,N3-derivative. Higher yield for the N3-benzoyl-
thymine (2) was obtained by one step reaction, extending the
treatment of thymine with benzoyl-chloride at 72–78 h.
The Mitsunobu reaction between 1 and 2 has been carried out
by using triphenylphosphine, di-tert-butyl-azodicarboxylate as
coupling agents and dioxane as solvent. The condensation product
N3-benzoyl-N1-[(2,2,5-trimethyl-1,3-dioxan-5-yl)methyl]-thymine
(3) was then treated with DowexÒ 50W-X8(H+) resin33 obtaining
N3-benzoyl-N1-[3-hydroxy-2-(hydroxymethyl)-2-methylpropyl]-
thymine (4). The latter derivative has been characterized by mono-
structure–activity relationships, which have been, until today, re-
ferred to the TBA-thrombin complex models fulfilled by NMR12–
14
and/or crystal11,15 methods, in which the aptamer adopts the
chair-like structure.
In light of literature data and with the aim to get new insight
about the thrombin-TBA recognition, we have undertaken a study
concerning the synthesis and the structural and biological charac-
terization of a new series of modified TBA, in which the thymidines
have been replaced, one at the time, by the acyclic nucleoside
N1-(3-hydroxy-2-hydroxymethyl-2-methylpropyl)-thymine
(T*,
Fig. 2). Various modifications involving the loops have been
brought on the TBA sequence, showing that, besides the G-quar-
tets, they are involved in the recognition between thrombin and
TBA.22–25 Oligonucleotides (ONs) incorporating acyclic nucleotides
are of current interest for many reasons. For example they are less
susceptible to nuclease action in relation to natural counterpart.26–
28
The increasing interest toward these derivatives is also due to
the hypothesis that oligonucleotides characterized by the replace-
ment of all the sugar moieties by acyclic linkers could have repre-
sented the precursors of nucleic acids in probiotic times.29 For
these reasons a highest number of ONs containing acyclic nucleo-
tide derivatives has been synthesized. In most of the cases a carbon
atom of the ribose ring is lacking, thus achieving a structural sim-
plification of the resulting molecules.
The structural changes between the acyclic and the unmodified
T residue (Fig. 2) can be summarized as follows: (i) one bond short-
ening between T base and one phosphate group; (ii) the presence
of an hydrophobic moiety, namely a methyl group; (iii) the pres-
ence of a prochiral carbon in the linker that develops into a chiral
center when the modified nucleotide is inserted in the sequences.
The ability of the synthesized sequences to assume a quadruplex
conformation has been valued by NMR, UV, and CD spectroscopies.
O
O
O
N
O
O
O
CH2OH
N
O
O
i
N
+
N
H
O
O
1
2
3
ii
O
O
O
N
O
O
O
N
O
O
N
N
N
iv
iii
H
H
N
O
N
O
N
O
O
O
ODMT
OH
ODMT
OH
OH
N
O
P
N
N(i-Pr)2
O
P
O
P
O
O
O
O
O
OCH2CH2CN
O
5'
3'
O
6
5
4
O
O
O
O
O
P
P
O
Scheme 1. Synthesis of acyclic nucleoside phosphoramidite. Reagents: (i) Mitsun-
obu condensation: 1 (9 mmol), 2 (4 mmol), PPh3 (6 mmol), DtBAD (6 mmol), and
dioxane (25 mL); (ii) 3 (10 mmol), DOWEX (30 mmol), and H2O/CH3OH (1/9 v/v,
50 mL); (iii) 4 (8 mmol), DMTCl (2 mmol), DMAP (0.4 mmol), and Py (32 mL); (iv) 5
(4 mmol), DIPEA (16 mmol), 2-cianoethyl-diisopropyl-chlorophosphoramidite
(2 mmol), DCM (35 mL).
O
O
Figure 2. Structural difference between the acyclic (left) and T (right) nucleotides.

8246
T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
and bi-dimensional NMR spectroscopy experiments (1H, 13C,
NOESY, and COSY) and mass spectrometry. The conversion of 4
in the phosphoramidite building block (6) has been achieved by
using the standard procedure34 (see Section 5). The monomer 6
was then successfully used for the synthesis of modified TBAs by
an automated DNA synthesizer. The acyclic nucleoside has been in-
serted, one at the time, as a T mimic in each position of the loops
(Table 1).
of the molar ellipticity on the concentration was due to end-to-end
interaction between two monomolecular quadruplexes, induced
by the cooling procedure, the CD experiments were repeated at
30 °C. Unchanged results, besides a small diminution of molar
ellipticity (Fig. 3c and inset d), have been checked confirming the
intermolecular folding for all the ONs sequences. Surprisingly,
invariability of molar ellipticity has been found using the physio-
logical PBS (one example is reported in the Fig. 4) or a similar so-
dium phosphate buffer totally KCl free (data not shown). These
data suggest that the TBA and its analogues opt for the inter- or in-
tra-molecular folding according to the nature of the cations present
in the solution. It is noteworthy that the physiological concentra-
tion of KCl present in PBS (2.7 mM) is insufficient to shift these
structures from mono- to bimolecular forms.
2.2. UV and CD experiments
The ability of all the TBA analogues to fold into an antiparallel
quadruplex arrangement was tested by CD analyses. For each CD
experiment, an ON concentration of 1 ꢀ 10ꢁ5 M was used. Two sets
of experiments were obtained using one of the following buffers:
100 mM KCl, 10 mM KH2PO4, pH 7.4 (K buffer); 10 mM NaH2PO4,
2.7 mM KCl, 137 mM NaCl, pH 7.4 (phosphate-buffered saline,
PBS). CD spectra have shown a profile very similar to that obtained
for the natural TBA (I, Table 1) with positive and negative maxima
at 295 and 265 nm, respectively, which are generally accepted as
diagnostic for an antiparallel quadruplex arrangement.35
In order to assign, by CD analysis, the molecularity of the struc-
tures formed by the modified ONs, we have monitored the ampli-
tude of CD positive band at 295 nm produced by each ON at
different concentrations: 1 ꢀ 10ꢁ4, 1 ꢀ 10ꢁ5, and 1 ꢀ 10ꢁ6 M. It is
well known that a monomolecular structure produces CD bands
whose amplitude, expressed as molar ellipticity, is independent
of the concentration.21 All CD experiments registered at 10 °C in
K buffer have shown a dependence of the molar ellipticity on the
concentration, thus indicating that the folding is intermolecular
for all the sequences (one example is reported in Fig. 3). However,
the UV thermal analyses performed in the same buffer (Fig. 5b)
have shown ambiguous profiles. For all species the experiments
seem to indicate that two transitions occur, the first around 25–
30 °C and the latter around 50 °C. To exclude that the dependence
To evaluate the thermal stability36–39 of the quadruplex folded
structures of I–VII (Table 1), UV melting curves have been regis-
tered using a temperature interval scan from 10 to 80 °C at fixed
wavelength (295 nm). Each experiment has been repeated at two
different scan rates of 0.1 and 0.5 °C/min (not shown) in both
PBS and K buffers (a and b, respectively, in Fig. 5). In all cases,
changing the scan rate, no differences have been evidenced. Fur-
thermore, all melting (from 10 to 80 °C) and folding (from 80 to
10 °C) profiles have not shown hysteresis phenomena.
The sequence-stability relationship data reported in Table 1 are
in agreement with the structural model in which T9, T4, and T13
are stacked on the adjacent G-quartet with the latter two base
paired to each other.12–14 In fact, these types of interactions can
be totally (as in the case of V) or partially (as in the cases of III
and VII) disrupted by the introduction of the acyclic nucleotide
at these positions. On the contrary, not unexpectedly, II and VI
have shown no lowering of Tm values, while for IV a slight increase
Table 1
UV melting temperatures of TBA and its analogues
ON
Sequences
Tm (K buffer)
Tm (PBS)
I
II
GGTTGGTGTGGTTGG
GGT3*TGGTGTGGTTGG
GGTT4*GGTGTGGTTGG
GGTTGGT7*GTGGTTGG
GGTTGGTGT9*GGTTGG
GGTTGGTGTGGT12*TGG
GGTTGGTGTGGTT13*GG
50
51
47
54
39
51
47
33
34
30
37
22
34
30
III
IV
V
VI
VII
Figure 4. CD spectra of VII (a) and I (inset b) registered using physiological PBS
(137 mM NaCl, 2.7 mM KCl, 10 mM NaH2PO4, and pH 7.4) at three ON concentra-
tions: 1 ꢀ 10ꢁ4 M (solid line); 1 ꢀ 10ꢁ5 M (dotted line); 1 ꢀ 10ꢁ6 M (dashed line).
ON concentration: 1.0 ꢀ 10ꢁ5 M; K buffer: 100 mM KCl, 10 mM KH2PO4, pH 7.4;
PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM NaH2PO4, pH 7.4.
Figure 3. CD spectra at 10 and 30 °C of VII (a and c, respectively) and I (insets b and d, respectively) registered using potassium phosphate as dissolving buffer (100 mM KCl,
10 mM KH2PO4, and pH 7.4) at three ON concentrations: 1 ꢀ 10ꢁ4 M (solid line); 1 ꢀ 10ꢁ5 M (dotted line); 1 ꢀ 10ꢁ6 M (dashed line).

T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
8247
Figure 5. Normalized melting profile of I, III, IV, VI: (a) physiological PBS and (b) potassium phosphate buffer. The scan rate is fixed at 0.1 °C/min. Tm values are taken as
minimum point of each curve derivative calculated using the corresponding function in the UV JASCO program. All Tm values are reported in Table 1.
in Tm value was observed, thus suggesting a marginal role of resi-
dues T3, T12, and T7 on the structural stability.
utable to H8 and H6 protons of guanine and thymine bases, respec-
tively, are present.12–14 By increasing the temperature the intensity
of imino signals progressively decreases and, at the same time, the
number of the aromatic resonances doubles due to the contempo-
raneous presence in solution of the two mutual exchanging folded
and unfolded ONs (data not shown). Rising the temperature above
50 °C the disappearing of all imino protons signals takes place, and
the number of aromatic signals drops back again to fifteen due to
the exclusive presence in solution of the unfolded ON. Very similar
signal patterns (for some examples see Fig. 8), as well as similar
thermal behaviors, can be observed for the NMR spectra of all
but V the modified TBAs, thus suggesting a comparable folding.
In some cases, however, the 1H NMR spectra appear more crowded
due to the splitting of several proton signals. This finding
can be confidently explained by considering that the substitution
Finally, for all ONs, differential UV (Fig. 6) and CD spectra
(Fig. 7) recorded at different temperatures have been carried out
using physiological PBS. In all cases but V, the data evidenced isoel-
liptic and isosbestic points, thus indicating that only one typology
of monomolecular quadruplex is present in solution.21
2.3. 1H NMR experiments
The 1H NMR spectrum of TBA is characterized by the presence
of eight well-resolved signals in the 11.5–12.5 ppm region attribut-
able to the exchange protected imino protons involved in the for-
mation of Hoogsteen hydrogen bonds of two G tetrads. Moreover,
in the aromatic region (6.8–8.5 ppm) fifteen proton signals attrib-
Figure 6. Differential UV spectra of VII (a) and I (b). Each differential profile is obtained by the subtraction of the spectrum registered at 65 °C from that obtained at each
specific temperature. (10, 15, 20, 25, 30, and 35 °C). ON concentration: 1.0 ꢀ 10ꢁ5 M in physiological PBS.
Figure 7. CD spectra of VII (a) and I (b) registered at various temperatures. ON concentration: 1.0 ꢀ 10ꢁ5 M in physiological PBS.

8248
T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
Figure 8. 1H NMR imino (left) and aromatic (right) regions of II and VI (700 MHz, 2 °C).
of a T with the T* nucleoside produces two diastereomeric quadru-
plex-forming ONs. This effect is clearly observable in the upfield
region of the 1H NMR spectrum of some of the modified TBAs. In
the 1H NMR spectrum of I the methyl groups of thymines T4 and
T13 resonate far away (0.98 and 1.03 ppm, respectively) from
those of the other thymines (ꢂ 2.0 ppm), due to their particular po-
sition in the quadruplex folded structure (see later). Given that in
all the modified TBAs reported here the aliphatic methyl group of
the acyclic linker has been introduced, a total of three 3 H singlets
were expected around 1.0 ppm. In almost all cases, however, more
than three signals appear in that region. Particularly, for III and VII
(Fig. 9) where the acyclic nucleoside substitutes T4 or T13, respec-
tively, five and six signals, respectively, are clearly discernible. It is
to be noted that by increasing the temperature above 45 °C the T4
and T13 methyl signals are downfield shifted, like those of the
other thymines, whereas the linker methyl group resonates as a
slightly split couple of singlets centered at 0.78 ppm as a conse-
quence of the presence of the two diastereomeric random coil
strands (data not shown).
of fibrinogen to the thrombin.6,7 Differently from other thrombin
inhibitors that bind to the catalytic site, the TBA binds the anionic
exosite of the thrombin leaving unaltered the thrombin ability to
cleave small chromogenic substrates. For this reason the inhibition
constant (Ki) of TBA analogues cannot be determined by traditional
methods, in which small chromogenic substrates are generally
used. In these cases, the most commonly used methods to evaluate
the anti-thrombin activity are the prothrombin time6,20,23 (PT) (or
INR or Quick time) assay and purified fibrinogen clotting assay.6,24
The PT assay is a routine diagnostic assay that evaluates in vitro
the activation of extrinsic pathway of the coagulation cascade. It
measures the time to clot upon the addition of excess of tissue fac-
tors to the plasma. The assay is most sensitive to the levels of the
extrinsic pathway factor VII and ‘‘common” pathway factors I
(fibrinogen), II (prothrombin), V, and X. The thromboplastin, the re-
agent used in the assay, consists of tissue factors and a mixture of
phospholipids and calcium. Briefly, the thromboplastin induces a
rapid conversion of plasmatic prothrombin into thrombin, which,
in turn, converts fibrinogen into fibrin, forming the clot. In absence
of a thrombin inhibitor, the time that elapses from the addition of
thromboplastin to the formation of the clot represents the basal
prothrombin time. When the assay is performed in presence of
the unmodified aptamer I the binding of fibrinogen to the throm-
bin is inhibited and more time is required to form the clot. The
TBAs analogues that prolong PT time more than unmodified I can
be taken into account as better thrombin inhibitors.
In contrast to the other modified ONs, the 1H NMR spectrum of
V is very crowded, thus confirming that several conformations are
simultaneously present in solution as suggested by the CD and UV
experiments in which isoelliptical and isosbestic points are miss-
ing (not shown).
2.4. Prothrombin Time (PT) assay
Prothrombin times have been determined by using an auto-
mated coagulometer. To perform this assay we have used fresh cit-
rated human plasma, thromboplastin, and a known concentration
It has been demonstrated that the unmodified I acts as a potent
anticoagulant as a consequence of its ability to prevent the binding
Figure 9. Upfield region of 1H NMR spectra of III and VII showing the split of linker, T4, and T13 methyl signals (500 MHz, 25 °C).

T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
8249
of the ONs. The plasma has been incubated with 2 or 20
l
M solu-
37 °C for few minutes. The time required to form the fibrin clot
has been measured by using a coagulometer. When TBA was
added to the fibrinogen buffer solution, the clotting time was pro-
longed in a concentration-dependent manner. In order to com-
pare the inhibitory activity of I–VII we have determined, for
each of them, the concentration required to double the basal clot-
ting time (Table 3).
tions of each ON at 37 °C. A set of experiments has been performed
using different incubation times: 30 s, 2, 5, 10, 15, 30, and 60 min.
After each incubation time, the clotting was initiated by addition of
thromboplastin and the PT time was measured by analyzing the
test on the coagulometer. Each PT measurement was produced in
triplicate, so that, for each incubation times, an averaged PT time
was calculated. Finally, for each ON, all the PT time values, mea-
sured at different incubation time, have been averaged in order
to furnish the final PT value and its standard deviation. The results
for unmodified I, reported in the Table 2, overlap with those previ-
ously obtained.25
Despite the experimental conditions used here were the same
previously reported6,24
, the clotting time value obtained for
unmodified I was about tenfold lower, whereas the basal clotting
time perfectly overlapped with the previously reported findings.
As shown in Table 3, IV, VI, and VII double the clotting time at low-
er concentration than I, whereas II, III, and V require a higher con-
centration. These results are comparable to those obtained by PT
assay, implying that the in vitro activities of these ONs are directly
associated to their thrombin binding ability.
The analysis of PT assays suggests that, although all modified
TBAs have shown anti-thrombin activity at 20
lM, only IV, VI,
and VII were significantly active at 2 M (Table 2). The highest
l
PT value was obtained for ON IV (Fig. 10) in which the modified
nucleoside occupies the position 7 in the 15-mer sequence. The
lowest PT values were obtained for ONs II, III, and V in which
the modified nucleoside occupies the positions 3, 4, and 9, respec-
tively. Interestingly the modified V, that shows the lowest thermal
stability and the most complex 1H NMR spectrum, induces a PT
time similar to that observed for II and III.
3. Discussion
The TBA (I) and its analogues (II–VII) show a similar behavior as
far as the folding typology and the biological properties are con-
cerned. As concerning CD analyses, the independence of molar
ellipticity from the concentration has indicated that the folding is
intramolecular when PBS is used as dissolving buffer. On the other
hand, experiments performed in potassium buffer have shown that
the molar ellipticity depends upon the concentration for all ONs, at
both 10 and 30 °C. Vorlikôva and co-workers21 have reported that
the solution structure of TBA aptamer is bimolecular. However, our
CD analyses confirm this datum only in potassium buffer, suggest-
ing that the molecularity of the folded structures could be affected
by the type of cations present in solution. Taking into account that
2.5. Purified fibrinogen clotting time
The clotting of fibrinogen had been previously used to valuate
the ability of TBA to bind thrombin and to inhibit the fibrinogen
polimerization.6,24 This assay is very similar to that of PT, but by
avoiding the interaction among ONs and other blood components,
it can be used to ascertain the direct relationship between the
in vitro properties of the modified ONs, evaluated by PT assay,
and their thrombin binding ability. To measure the clotting time
in absence of any inhibitor (i.e., the basal clotting time), thrombin
was added to a fibrinogen buffer solution, previously incubated at
Table 3
Concentration required for each ON to double the fibrinogen clotting time
ON
[ON]2t nM ( 10)
Table 2
I
II
175
238
250
70
210
78
Concentration-dependent response, following the ONs incubation (up to 60 min at 2
or 20
l
M) with human plasma, on PT value expressed in seconds
III
IV
V
VI
VII
ON
PT (2 M)
l
PT (20 lM)
I
II
22.60 1.83
14.30 0.31
14.50 0.20
24.67 0.59
13.23 0.20
18.07 0.47
17.50 0.23
53.68 1.48
35.15 1.70
32.28 0.21
60.20 1.38
34.65 2.01
44.83 1.30
46.97 1.83
100
III
IV
V
VI
VII
The basal time is determined by measuring the time required to form the clot in
200 M of a fibrinogen buffer solution (2 mg in 1 mL of 20 mM Tris–acetate buffer
containing 140 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, and 1 mM CaCl2, pH 7.4) after
the addition of 100 M of buffer thrombin solution (at final concentration of
13 nM). Basal clotting time, averaged on six measurements, has been 27 3 s. The
ON concentration required to prolong the clotting time at final value of 57 10 s
was determined (see Section 5).
l
l
The physiological value of human PT is 13.4 0.22 s. PBS, used as vehicle, does not
alter PT value (12.49 0.39).
70
30
°°°
***
60
***
***
***
50
***
***
20
***
***
40
30
20
10
0
10
0
2 µM
20 µM
***
Figure 10. Incubation, up to 15 min, of I, IV, VI, and VII significantly and concentration-dependently increases the PT value. p < 0.001 vs vehicle, °°°p < 0.001 vs I. n = 3.

8250
T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
PBS mimes the blood salt composition, the intermolecular folding
for which a highest potassium concentration is required should
not be relevant from biological point of view.
As far as modifications involving the TT loops are concerned, no
significant effect on quadruplex stability has been shown when the
substitution concerns the positions 3 or 12. On the contrary, mod-
ified TBAs in which the substitution involves the position 4 or 13
have shown a lowering of the thermal stability. These results can
be explained considering the NMR structure12,13,20 in which T4
and T13 stack on adjacent G quartet and bind each other by a
hydrogen bond. The introduction of an acyclic nucleotide at one
of these two positions can induce the loss of stacking interactions
and/or the loss of the H-bond, hence the diminution of Tm value.
A detailed analysis of the structure–activity relationship results
in more difficulty because the X-ray structure of TBA–thrombin
complex shows poor resolution about the TT loops posi-
tions.11,12,15,20 As a consequence, the results obtained for II or VII,
in which the acyclic nucleotide occupies the positions 3 and 13,
respectively, were particularly intriguing. In fact, VII folds in a
quadruplex structure having poor thermal stability, but among
ONs in which TT loops have been modified; it shows the highest
anti-thrombin activity. On the other hand, although II has kept
the original thermal stability, it has partially lost the biological
activity.
In order to shed light on our results and considering the tem-
perature of the PT assay, the presence of an equilibrium between
the folded form and the unfolded form, both likely being able to
bind the thrombin, must be taken into account. In fact, Baldrich
et al.40 and Nagatoishi et al.41 have demonstrated that thrombin
is able to act as a molecular chaperon for the unfolded form of
the TBA thus inducing the G-quadruplex formation. Furthermore,
authors hypothesize that the recognition by ‘‘adaptive folding”
could start by an interaction between the TT steps of the unfolded
TBA and some hydrophobic residues of the protein. In this context,
the recognition of the unfolded form by the protein could contrib-
ute significantly to the biological activity. Results concerning the
PT assay (Table 2) and purified fibrinogen clotting assay (Table 3)
obtained for ONs, in which TT loops have been modified, clearly
indicate that modifications involving T3 or T4 result in lower bio-
logical activities than those concerning T12 or T13. These data
seem to suggest that the chaperon process involves mainly the ini-
tial contact between thrombin and the TT loop flanking the 50-end.
The biological properties of V are comparable to that of II and III. It
should take into account that its ability to fold into a chair-like
quadruplex structure is dramatically compromised by the modifi-
cation, as evidenced by its thermal stability being far lower than
that of the other ONs.
The introduction of the acyclic nucleotide generates two diaste-
reomeric strands in which the modified residue may assume an R
or S configuration. Consequently, in PBS, the folding process of
each modified ON produces two diastereomeric structures. How-
ever, the presence of isoelliptical and isosbestic points in thermal
CD and differential UV spectra indicates that the two possible
structures show similar arrangement and thermal stability.
The 1H NMR spectra of all but V the modified ONs are similar to
that of the unmodified one, besides the splitting of some imino and
methyl signals. The presence of doubled signals in some regions
could be confidentially attributable to the two diastereomeric
structures.
The reported 1H NMR of the TBA12,13 showed in the upfield re-
gion two methyl signals at anomalous values (0.98 and 1.07 ppm,
respectively) due to the methyl groups of T4 and T13 bases. The
upfield shifts of these signals have been justified by the formation
of a base pair stacked on the adjacent G-quartet in the monomolec-
ular chair-like structure. It is noteworthy that in almost all cases a
similar upfield shift for T4 and T13 methyl groups has been ob-
served, thus suggesting that the structural arrangement is similar
to that of TBA. The 1H NMR of V appeared very crowded due to
the presence of more than one structure in solution.
The in vitro biological properties of the new ONs have been val-
ued by PT assay performed at 37 °C. Prior to the assay, each ON was
annealed using the physiological PBS. In order to properly get in-
sight into the structure-activity relationship, the spectroscopic
experiments have been carried out in the same buffer. The UV
and CD thermal experiments have shown that the Tm values of
TBA and its analogues were lower or comparable to the tempera-
ture of the PT assay, pointing out that in physiological conditions
they are present in solution mainly as unfolded ONs.
NMR and X-ray studies unequivocally assign to the TBA mono-
molecular quadruplex its ability to bind the thrombin with high
specificity.15,20 TBA-thrombin complex analyses evidence that the
G-scaffold is fundamental for the complex stability.11–15,20 As a di-
rect consequence, it is suited to hold that every modification pro-
duced on the ON sequence (aimed at the improvement of specific
properties such as the cellular permeability or the in vivo half-life)
can be considered appropriate only if the resulting ONs improve or
preserve the thermal stability of the folded chair-like quadruplex
structure. However, our experimental data only partially agree
with this overview, since a direct correlation between the struc-
tural stability and the biological activity is preserved mainly when
the introduction of the modified nucleotide involves the TGT loop,
whereas a more complex structure–activity relationship results in
the case of TT loops modifications. Similar behavior was also found
for other biologically active quadruplex-forming ON.35
Furthermore, more recently, it has been reported that TBA is
able to bind both thrombin and prothrombin with similar affin-
ity.42 This means that any variation introduced in the sequence
could interfere with the TBA ability to bind one or both the pro-
teins. Prothrombin is an attractive target for novel anticoagulant
agents since its inactivation determines the attenuation of throm-
bin generation preserving at the same time sufficient thrombin
activity to permit hemostasis. For this reason, the binding affinity
of modified ONs toward prothrombin should be evaluated for a
more detailed valuation of in vitro effects. These studies will be
successively performed by using the diastereomeric pure oligonu-
cleotides whose synthesis is currently in progress in our
laboratory.
Previous studies13,24 had demonstrated that the substitution of
the T7 with an abasic linker is associated with enhanced thrombin
binding ability, whereas when the abasic linker replaces the T9 a
loss of binding ability is observed. This is in agreement with the
analyses of the structural data of the TBA-thrombin complex,15
which have shown that the stacking of T9 and G8 on the adjacent
G-quartet is important for the complex stability.
The same behavior has been observed for ONs IV and V in which
the modification involves T7 and T9, respectively, thus confirming
that the T9 and G8 stacking interactions are both required for the
quadruplex stability and the thrombin binding activity. The
enhancing of biological activity obtained for IV and other abasic
site containing ONs24 indicates that improved thrombin binding
activity is achieved by increasing the flexibility at position 7 of
TBA. This datum should be taken into account for the design of
new TBA analogues.
4. Conclusion
In summary, the synthesis of a series of modified TBA analogues
in which the T bases have been replaced, one at the time, by the
new acyclic nucleotide N1-(3-hydroxy-2-hydroxymethyl-2-meth-
ylpropyl)-thymine has been done. Each ON has been characterized
by UV, CD, MS, and 1H NMR experiments. The structural data have
demonstrated that all ONs fold into quadruple helices whose

T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
8251
behavior is very similar to that of the unmodified counterpart. The
spectroscopic data suggest that the molecularity of the formed
quadruplex structure depends upon the nature of the cations pres-
ent in the buffer. The biological activity of the new ONs has been
valued by prothrombin time measurements and by purified fibrin-
ogen clotting experiments. The modified ON containing the acyclic
nucleotide at T7 position has shown a higher in vitro inhibitory
activity than the unmodified sequence at the two tested concentra-
sets were zero filled to 32,768 points prior to Fourier transforma-
tion and apodized with a shifted sine bell squared window func-
tion. Pulsed-field gradient DPFGSE43,44 sequence was used for
H2O suppression. NMR samples II–VII were prepared in H2O/D2O
(9:1, v/v) at a single strand concentration of approximately
0.5 mM, with a final salt concentration of 130 mM NaCl, 10 mM
NaH2PO4, and 0.2 mM EDTA.
tions of 2 and 20
lM.
5.4. Prothrombin time (PT) assay
Some new structure–activity relationships have been evidenced
when the modification has involved the TT loops. Particularly, we
suggest that important interactions must involve the TT loop lo-
cated at 50-end, most probably during the recognition of the un-
folded form of TBA and the other ONs by the thrombin. At the
best of our knowledge data concerning the events involved in the
chaperon process have not been presented as yet. Since the TBA
can be considered one of the most promising anticoagulant drugs,
the full elucidation of its structure–activity relationships is partic-
ularly interesting. These results can contribute to the understand-
ing of TBA-thrombin recognition process.
Human plasma samples were collected by venipuncture in pres-
ence of 0.1 volumes of 3.8% sodium citrate and fractionated by cen-
trifugation at 2000 rpm for 5 min. PT time was measured by using
Koagulab MJ Coagulation System with a specific kit RecombiPlas
Tin HemosIL (Inst. Labs, Lexinton, USA). Briefly, this method relies
on the high sensitivity thromboplastin reagent based on recombi-
nant human tissue factors. The addiction of recombiplastin to the
plasma in presence of calcium ions initiates the activation of
extrinsic pathway converting the fibrinogen into fibrin, with a for-
mation of solid gel. The procedure was performed according to the
manufacturer’s instructions. In our experimental protocol a time
course of each ON or vehicle incubation with 100
37 °C has been performed. For the valuation of PT at 20
l
L of plasma at
M, in
5. Experimental
l
the apposite microtube, 2 lL of the corresponding ON solution
5.1. UV spectroscopic temperature-dependent melting studies
(1.0 ꢀ 10ꢁ3 M in PBS) or vehicle was added. After 30 s, 1, 5, 15,
Each ON, at a final concentration of 1.0 ꢀ 10ꢁ5 M, was dissolved
in the potassium (100 mM KCl, 10 mM KH2PO4, pH 7.0) or sodium
(PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM NaH2PO4, pH 7.4) phos-
phate buffer.
All samples were submitted to the annealing procedure (heated
at 90 °C and slowly cooled at room temperature).
The UV thermal stability experiments were carried out using a
JASCO 530 spectrophotometer equipped with ETC-505 T tempera-
ture controller. Melting curves were obtained by monitoring the
variation of absorbance at 295 nm from 10 to 80 °C. Two melting
experiments for each ON were recorded, fixing the heating rate
at 0.5 and 0.1 °C/min, respectively. The folding experiments were
registered from 80 to 10 °C, using the same scan rates.
The differential absorption UV profiles were obtained by the
subtraction of the spectrum registered at 65 °C from that obtained
at each specific temperature (10, 15, 20, 25, 30, and 35 °C).
30, and 60 min of incubation, 200 lL of the kit solution containing
recombiplastin was added with consequent activation of extrinsic
pathway.
The PT at final ONs concentrations of 2 lM was determined
using 2
l
l of 1.0 ꢀ 10ꢁ4 M ON solution in PBS.
The PT measurement, for each incubation time, was produced in
triplicate and the average value was calculated. Finally, all the PT
time values measured at different incubation times were also aver-
aged to furnish the final PT value and its standard deviation.
5.5. Purified fibrinogen clotting assay
ONs were incubated for 1 min at 37 °C in 200 ll of buffer
(20 mM tris acetate, 140 mM NaCl, 2.7 mM KCl, 1 mM MgCl2,
1 mM CaCl2, pH 7.4) containing human fibrinogen (2 mg/mL).
The human thrombin was dissolved in the same buffer and pre-
equilibrated for 1 min at 37 °C. After this time 100 ll of thrombin
was added to the solution containing the fibrinogen and the ON
(the final concentration of thrombin was 13 nM). The time re-
quired to clot was measured using a Koagulab MJ Coagulation Sys-
tem. For each ON the experiment was performed at five different
concentrations (20, 50, 100, 175, 200, and 300 nM), and the clot-
ting time was determined by averaging six experiments for each
ON concentration. The basal clotting time has been determined
by measuring the clotting time in absence of any ONs. Clotting
time versus [ON] was then reported to obtain a straight line, from
which the concentration required to double the clotting, indicated
as [ON]2t, was obtained.
5.2. Circular dichroism spectra
The CD spectra were recorded from 200 to 360 nm (scanning
rate: 100 nm/min; response: 16 s; bandwidth: 1.0 nm) on a JASCO
715 spectropolarimeter equipped with a PTC-348 temperature
controller. All samples were previously submitted to the annealing
procedure. Before the scan, the sample was equilibrated at the spe-
cific temperature (10 or 30 °C) for 30 min. Each CD profile was ob-
tained by taking the average of three scans from which the
spectrum of the buffer alone was subtracted.
To evaluate the dependence of molar ellipticity from the ON con-
centration in PBS and potassium buffer, three samples at different
concentrations (1.0 ꢀ 10ꢁ4, 1.0 ꢀ 10ꢁ5, and 1.0 ꢀ 10ꢁ6 M) and three
cuvettes having different path lengths (0.1, 0.5, and 1.0 cm), were
used. The CD spectra at different temperatures have been registered
using samples at 1.0 ꢀ 10ꢁ5 M in PBS or potassium buffer.
5.6. Synthesis
5.6.1. Reagent and equipment
Chemicals and anhydrous solvents have been purchased from
Fluka–Sigma–Aldrich. TLCs were run on Merck silica gel 60 F254
plates, and compounds were visualized by KMnO4 spraying re-
agent, and UV lamp. Silica gel chromatographies were performed
by using Merck silica gel 60 (0.063–0.200 mm). API 2000 Analist
NT Applied Biosystem ESI-MS spectrometer was used for analyses
of the intermediates and the monomer. Reagents and phospho-
ramidites for DNA syntheses were purchased from Glen research.
ON syntheses were performed on a PerSeptive Biosystem Expedite.
5.3. NMR experiments
1H NMR data were collected on Varian Unity INOVA 500 and
700 MHz spectrometers equipped with broad-band inverse probes
with z-field gradient and processed using Varian VNMR software
package. 1D NMR spectra were acquired as 16,384 data points with
a recycle delay of 1.0 s at temperatures in the range 2–50 °C. Data

8252
T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
HPLC analyses and purifications were carried out by using Thermo-
Finnigan Spectra SYSTEM P4000. To perform the purifications of
the ONs a SAX 1000-8 Macherey–Nagel column was used. The
desalting of ONs was performed by using molecular exclusion
chromatography on BIORAD BioLogic LP system. To check the puri-
135.8, 132.1, 131.0, 129.9, 110.2, 65.8, 51.9, 42.0, 18.7, 13.3. ESI
MS m/z 333 [M+H]+, 355 [M+Na]+.
5.6.6. N3-Benzoyl-N1-[3-methoxy(4,40-dimethoxytrityl)-2-hydroxy-
methyl-2-methylpropyl]-thymine (5)
fication grade a PurospherÒ STAR RP-18 end-capped (5-
lm) HPLC
4 (1.54 mmol), 4,40-dimethoxytrytyl chloride (1.23 mmol), and
4-dimethylaminopyridine (0.077 mmol) were dissolved in dry pyr-
idine (10 mL) and dry acetonitrile (5 mL). The resulting solution
was stirred for 1 h at room temperature under argon. Dry methanol
was added to quench the reaction. After 30 min under stirring, the
solution was concentrated under reduced pressure and the residue
purified by column chromatography on silica gel eluted with
50:50:5 EtOAc/hexane/Et3N to give a clear-yellow solid 5 (yield
45%), Rf 0.51 (1:1 v/v EtOAc/hexane). 1H NMR (CDCl3) d ppm
7.90 (d, 2H), 7.65 (t, 1H), 7.48 (t, 2H), 7.40 (d, 2H), 7.38 (m, 3H),
7.35 (d, 4H), 7.12 (s, 1H), 6.85 (d, 4H), 3.95–3.75 (dd, 2H), 3.80 (s,
6H), 3.38–3.21 (m, 2H), 3.10–2.95 (m, 2H), 1.80 (s, 3H), 1.05 (s,
3H). 13C NMR (CDCl3) d ppm 167.7, 162.1, 158.8, 151.1, 143.6,
142.1, 135.0, 135.5, 132.1, 131.0, 129.9, 129.5, 129.2, 128.4,
126.7, 113.4, 110.2, 94.5, 71.8, 65.9, 55.4, 51.2, 42.0, 18.7, 13.3.
ESI MS m/z 635 [M+H]+, 657 [M+Na]+.
column was used. Mass spectrometric analyses of ONs were per-
formed on Bruker Autoflex I MALDI-TOF mass spectrometer using
piconilic/3-hydroxypiconilic acids mixture as the matrix.
5.6.2. 5-Hydroxymethyl-2,2,5-trimethyl-1,3-dioxane (1)
1,1,1-Tris(hydroxymethyl)ethane (41.63 mmol) was suspended
in 50 mL of 2,2-dimethoxypropane. Pyridinium-p-toluene sulfo-
nate (8.32 mmol) was added and the suspension was stirred at
room temperature for 24 h under argon. Triethylamine was added
to quench the reaction, and the solution was stirred for 30 min. The
solvent was removed to leave a colorless liquid. The solid residue
was chromatographed on silica gel eluted with 1:1 EtOAc/hexane
to give a colorless liquid 1 (yield 67%), Rf 0.40 (1:1 EtOAc/hexane).
1H NMR (CDCl3) d ppm 3.71–3.56 (dd, 4 H), 3.38 (s, 2H), 1.41 (s,
3H), 1.39 (s, 3H), 0.83 (s, 3H). 13C NMR (CDCl3) d ppm 90.1, 68.2,
65.9, 37.0, 28.3,13.3. ESI MS m/z 161 [M+H]+, 183 [M+Na]+.
5.6.7. 3-[3-Benzoyl-5-methyl-2,4-dioxo-3,4-dihydropyrimidin-
1(2H)-yl]-2-methoxy-(4,40-dimethoxytrity1)-2-methylpropyl-2-
cyanoetyl-diisopropylphosphoramidite (6)
5.6.3. N3-Benzoylthymine (2)
Thymine (8.55 mmol) was suspended in dry acetonitrile
(10 mL) and dry pyridine (5 mL), and the mixture was cooled at
0 °C. Benzoyl chloride (19.1 mmol) was added and the reaction
mixture was stirred at room temperature for 72 h, under argon.
The mixture was evaporated under reduced pressure, and the res-
idue was dissolved in dichloromethane and extracted with water
(twice). The product was crystallized from dichloromethane giving
colorless needles 2 (yield 85%), Rf 0.45 (9:1 CHCl3/MeOH). 1H NMR
(CDCl3) d ppm 9.57 (br, 1H), 7.95 (d, 2H), 7.68 (t, 1H), 7.50 (t, 2H),
7.12 (s, 1H), 1.95 (s, 3H). 13C NMR (CDCl3) d ppm 167.7, 162.1,
151.1, 142.1, 135.8, 132.1, 131.0, 129.9, 110.2, 18.7. 2D-NOESY
NMR (CDCl3) d ppm: a NOE cross peak was found between H-6
(7.12) and N1H (9.57). ESI MS m/z 231[M+H]+, 253 [M+Na]+.
5 (0.559 mmol) was dried in vacuo overnight before being dis-
solved in anhydrous CH2Cl2 (5 mL) and diisopropylethylamine
(2.24 mmol, 4 equiv) under argon. To this solution, 2-cyanoethyl
diisopropylchlorophosphoramidite was added (0.838 mmol, 1.5
equiv). After 20 min the reaction mixture was quenched by the
addition of dry methanol, diluted with ethyl acetate (15 mL) and
washed with 10% sodium carbonate solution (15 mL) and brine
(15 mL). The organic layer was dried on magnesium sulfate and
concentrated in vacuo. The residue was purified on silica gel col-
umn eluted with 80:10:10 v/v/v CH2Cl2:ethyl acetate:triethyl-
amine. The fractions containing the product were collected and
concentrated under vacuum yielding a white foam 6 (82%), Rf
0.65 (97:3 CHCl3/MeOH). 1H NMR (CDCl3) d ppm 7.90 (d, 2H),
7.65 (t, 1H), 7.48 (t, 2H), 7.40 (d, 2H), 7.38 (m, 3H), 7.35 (d, 4H),
7.12 (s, 1H), 6.85 (d, 4H), 3.95–3.80 (m, 2H), 3.75 (s, 6H), 3.70
(m, 2H), 3.60 (m, 2H), 3.55 (m, 2H) 3.39-3.02 (m, 2H), 2.60–2.45
(t, 2H), 1.80 (s, 3H), 1.09 (s, 12H) 1.05 (s, 3H). 13C NMR (CDCl3) d
ppm 163.3, 158.7, 150.6, 141.8, 135.0, 130.6, 130.4, 129.3, 129.2,
128.4, 128.1, 128.1, 127.0, 113.4, 110.1, 94.5, 66.1, 55.4, 54.2,
51.2, 43.3, 42.0, 24.7, 20.7, 18.7, 13.3. ESI MS m/z 835.7 [M+H]+,
857.7 [M+Na]+.
5.6.4. 3-Benzoyl-5-methyl-1-[(2,2,5-trimethyl-1,3-dioxan-5-yl)-
methyl]-pyrimidine-2,4(1H,3H)-dione (3)
2 (4.0 mmol), triphenylphosphine (6.0 mmol), di-tert-butyl azo-
dicarboxylate (6.0 mmol) were suspended in 25 mL of dry dioxane;
the resultant mixture was stirred and cooled to ꢁ20 °C and 1 was
added. The reaction mixture was stirred at room temperature for
18 h under argon and a clear-yellow solution was obtained. The
solution was concentrated under reduced pressure and the residue
was purified by column chromatography on silica gel eluted with
95:5 Et2O/CH2Cl2 to give a white solid 3 (yield 75%), Rf 0.59 (9:1
v/v CHCl3/MeOH). 1H NMR (CDCl3) d ppm 7.92 (d, 2H), 7.63 (t,
1H), 7.48 (t, 2H), 7.19 (s, 1H), 4.02 (s, 2H), 3.71–3.57 (dd, 4H),
1.97 (s, 3H), 1.47 (s, 3H), 1.46 (s, 3H), 0.79 (s, 3H). 13C NMR (CDCl3)
d ppm 167.7, 162.1, 151.1, 142.1, 135.8, 132.1, 131.0, 129.9, 110.2,
90.1, 67.0, 51.9, 37.0, 28.3, 18.7, 13.3. ESI MS m/z 373 [M+H]+, 395
[M+Na]+.
5.7. Synthesis of oligomers
TBA and analogues were synthesized by using standard solid-
phase DNA chemistry on controlled pore glass (CPG) support fol-
lowing the b-cyanoethyl phosphoroamidite method.28 The oligo-
mers were detached from the support and deprotected by
treatment with an aqueous ammonia solution (33%) at 55 °C over-
night. The combined filtrates and washings were concentrated un-
der reduced pressure, dissolved in H2O, and purified by HPLC using
an anionic exchange column eluted with a linear gradient (from 0%
to 100% B in 30 min) of phosphate buffer at pH 7.0 (A: 20 mM
NaH2PO4 aqueous solution containing 20% CH3CN; B: 1 M NaCl,
20 mM NaH2PO4 aqueous solution containing 20% CH3CN). The
oligomers were successively desalted by molecular exclusion chro-
matography on Biogel P-2 fine. The purity was checked on HPLC by
using reverse phase column (C-18 Purospher STAR, Merck) and
electrophoresis on 20% denaturing polyacrylamide gel containing
7 M urea. The concentrations of the samples used in CD and UV
experiments were determined by measuring absorbance at
5.6.5. N3-Benzoyl-N1-(3-hydroxy-2-hydroxymethyl-2-methylpropyl)-
thymine (4)
3 (2.83 mmol) was suspended in 9:1 MeOH/H2O (50 mL), and
3 g of DowexÒ 50WX8 (H+) resin was added. The reaction mixture
was slowly stirred at room temperature for 8 h up to complete dis-
solution of 3 and then a 0.5 M aqueous NaOH solution was added
up to neutralization. The solution was concentrated under reduced
pressure, obtaining 4 as a white solid (yield 98%); Rf 0.27 (9:1
CHCl3/MeOH). 1H NMR (CDCl3) d ppm 7.98 (d, 2H), 7.70 (t, 1H),
7.60 (s, 1H), 7.57 (t, 2H), 3.80 (s, 2H), 3.20 (m, 4H), 1.95 (s, 3H),
0.89 (s, 3H). 13C NMR (CDCl3) d ppm 167.7, 162.1, 151.1, 142.1,

T. Coppola et al. / Bioorg. Med. Chem. 16 (2008) 8244–8253
8253
16. Tuerk, C.; Gold, L. Science 1990, 249, 505–510.
17. Abelson, J. Science 1990, 249, 488–489.
260 nm at 80 °C using the extinction coefficients calculated accord-
ing to Gray et al. method.45 The extinction coefficient of the mod-
ified acyclic nucleotide has been assumed equal to that of the
unmodified thymidine.
18. Nimjee, S. M.; Rusconi, C. P.; Sullenger, B. A. Annu. Rev. Med. 2005, 56, 555–583.
19. Mao, X.; Gmeiner, W. H. Biophys. Chem. 2005, 113, 155–160.
20. Kelly, J. A.; Feigon, J.; Yeates, T. O. J. Mol. Biol. 1996, 256, 417–422.
21. Fialová, M.; Kypr, J.; Vorlìcková, M. Biochem. Biophys. Res. Commun. 2006, 344,
50–54.
22. He, G.-X.; Krawczyk, S. H.; Swaminathan, S.; Shea, R. G.; Dougherty, J. P.;
Terhorst, T.; Law, V. S.; Griffin, L. C.; Coutré, S.; Bischofberger, N. J. Med. Chem.
1998, 41, 2234–2242.
MALDI-TOF MS m/z (calcd. 4738 [MꢁH]ꢁ, 4761 [Mꢁ2 H+Na]ꢁ)
4736 (II, [MꢁH]ꢁ), 4736 (III, [MꢁH]ꢁ), 4735 (IV, [MꢁH]ꢁ), 4734
(V, [MꢁH]ꢁ), 4763 (VI, [Mꢁ2H+Na]ꢁ), 4762 (VII, [Mꢁ2H+Na]ꢁ).
23. He, G.-X.; Williams, J. P.; Postich, M. J.; Swaminathan, S.; Shea, R. G.; Terhorst,
T.; Law, V. S.; Mao, C. T.; Sueoka, C.; Coutré, S.; Bischofberger, N. J. Med. Chem.
1998, 41, 4224–4231.
Acknowledgements
24. Krawczyk, S. H.; Bischofberger, N.; Griffin, L. C.; Law, V. S.; Shea, R. G.;
Swaminathan, S. Nucleosides Nucleotides Nucleic Acids 1995, 14, 1109–1116.
25. Martino, L.; Virno, A.; Randazzo, A.; Virgilio, A.; Esposito, V.; Giancola, C.; Bucci,
M.; Cirino, G.; Mayol, L. Nucleic Acids Res. 2006, 34, 6653–6662.
26. Merle, Y.; Bonneil, E.; Merle, L.; Sage, J.; Szmezö, A. Int. J. Biol. Macromol. 1995,
17, 239–246.
27. Marshalko, S. J.; Schweitzer, I. B.; Beardsley, G. P. Biochemistry 1995, 34, 9235–
9248.
28. Zhang, L.; Periz, A.; Meggers, E. V. J. Am. Chem. Soc. 2005, 127, 4174–4175.
29. Joyce, G. F. Nature 2002, 418, 214–221.
This work was financially supported by Regione Campania
(Legge 5) and M.I.U.R. (Prin). The authors thank Luisa Cuorvo for
technical assistance.
References and notes
1. Bode, W. Blood Cell. Mol. Dis. 2006, 36, 122–130.
2. Dahlback, B. J. Int. Med. 2005, 257, 209–223.
3. Colman, R. W.; Hirsh, J.; Marder, V. J.; Clowes, A. W. In Hemostasis and
Thrombosis: Basic Principles and Clinical Practice, 4th ed.; Lippincott Williams
and Wilkins: Philadelphia, 2004; pp 17–20.
30. Mitsunobu, O. Synthesis 1981, 1, 1–28.
31. Di Grandi, M. J.; Tilley, J. W. Tetrahedron Lett. 1996, 37, 4327–4330.
32. Cruickshank, K. A.; Jiricny, J.; Reese, C. B. Tetrahedron Lett. 1984, 25,
681–684.
33. Park, K. H.; Yoon, Y. J.; Lee, S. G. Tetrahedron Lett. 1994, 35, 9737–9740.
34. Sinha, N. D.; Biernat, J.; McManus, J.; Koster, H. Nucleic Acids Res. 1984, 12,
4539–4557.
4. Weitz, J. I.; Hirsh, J. Chest 2001, 119, 95S–107S.
5. Di Nisio, M.; Middeldorp, S.; Buller, H. R. N. Engl. J. Med. 2005, 353, 1028–1040.
6. Bock, L. C.; Griffin, L. C.; Latham, J. A.; Vermaas, E. H.; Toole, J. J. Nature 1992,
355, 564–566.
7. Griffin, L. C.; Tidmarsh, G. F.; Bock, L. C.; Toole, J. J.; Leung, L. L. K. Blood 1993, 81,
3271–3276.
8. Tsiang, M.; Gibbs, C. S.; Griffin, L. C.; Dunn, K. E.; Leung, L. L. K. J. Biol. Chem.
1995, 270, 19370–19376.
9. Paborsky, L. R.; McCurdy, S. N.; Griffin, L. C.; Toole, J. J.; Leung, L. L. K. J. Biol.
Chem. 1993, 268, 20808–20811.
10. Tasset, D. M.; Kubik, M. F.; Steiner, W. J. Mol. Biol. 1997, 272, 688–698.
11. Padmanabhan, K.; Padmanabhan, K. P.; Ferrara, J. D.; Sadler, J. E.; Tulinsky, A. J.
Biol. Chem. 1993, 268, 17651–17654.
12. Macaya, R. F.; Schultze, P.; Smith, F. W.; Roe, J. A.; Feigon, J. Proc. Natl. Acad. Sci.
U.S.A 1993, 90, 3745–3749.
13. Wang, K. Y.; McCurdy, S.; Shea, R. G.; Swaminathan, S.; Bolton, P. H.
Biochemistry 1993, 32, 1899–1904.
14. Wang, K. Y.; Krawczyk, S. H.; Bischofberger, N.; Swaminathan, S.; Bolton, P. H.
Biochemistry 1993, 32, 11285–11292.
´
´
35. Dapic, V.; Abdomerovic, V.; Marrington, R.; Peberdy, J.; Rodger, A.; Trent, J. O.;
Bates, P. J. Nucleic Acids Res. 2003, 31, 2097–2107.
36. Mergny, J.-L.; Phan, A.-T.; Lacroix, L. FEBS Lett. 1998, 435, 74–78.
37. Mergny, J.-L.; Li, J.; Lacroix, L.; Amrane, S.; Chaires, J. B. Nucleic Acids Res. 2005,
33, e138/1–e138/6.
38. Saccà, B.; Lacroix, L.; Mergny, J.-L. Nucleic Acids Res. 2005, 33, 1182–1192.
39. Gros, J.; Rosu, F.; Amrane, S.; De Cian, A.; Gabelica, V.; Lacroix, L.; Mergny, J.-L.
Nucleic Acids Res. 2007, 35, 3064–3075.
40. Baldrich, E.; O’Sullivan, C. K. Anal. Biochem. 2005, 341, 194–197.
41. Nagatoishi, S.; Tanaka, Y.; Tsumoto, K. Biochem. Biophys. Res. Commun. 2007,
352, 812–817.
42. Kretz, C. A.; Stafford, A. R.; Fredenburgh, J. C.; Weitz, J. J. Biol. Chem. 2006, 281,
37447–37485.
43. Hwang, T. L.; Shaha, A. J. J. Magn. Res. 1995, A112, 275–279.
44. Dalvit, C. J. Biomol. NMR 1998, 11, 437–444.
45. Gray, D. M.; Hung, S. -H.; Johnson, K. H. Methods Enzymol. 1995, 246, 19–34.
15. Padmanabhan, K.; Tulinsky, A. Acta Crystallogr. 1996, D52, 272–282.