Investigations toward the selection of fully-modified 4′-thioRNA aptamers: Optimization of in vitro transcription steps in the presence of 4′-thioNTPs

Article abstract of DOI:10.1016/j.bmc.2008.09.048

We describe herein a method for isolating fully-modified 4′-thioRNA aptamers against human α-thrombin using the SELEX protocol. In order to isolate the desired aptamers, in vitro transcription was examined in the presence of four kinds of 4′-thioribonucleoside triphosphates (4′-thioNTPs) in an effort to afford the fully-modified 4′-thioRNA. The transcription efficiency of the 4′-thioNTPs was first compared with that of the nucleoside 5′-(α-thio)triphosphates (αSNTPs) and found to be less effective than that of the αSNTPs, especially when GTP and/or ATP were substituted for 4′-thioGTP and/or 4′-thioATP. Further attempts to improve its efficiency, including the use of a mutant RNA polymerase instead of the wild type, various additives, and 4′-thioNTP concentrations were unsuccessful. Accordingly, the transcription was performed in the presence of 4′-thioNTPs together with the natural GTP and ATP at the appropriate concentrations. Although this attempt furnished a highly-modified 4′-thioRNA, but not a fully-modified 4′-thioRNA, we eventually succeeded in isolating the fully-modified 4′-thioRNA aptamers by SELEX using optimized transcription conditions, followed by post-modification of the resulting aptamers.

Full text of DOI:10.1016/j.bmc.2008.09.048

Bioorganic & Medicinal Chemistry 16 (2008) 9450–9456
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Investigations toward the selection of fully-modified 40-thioRNA aptamers:
Optimization of in vitro transcription steps in the presence of 40-thioNTPs
*
*
Noriaki Minakawa , Mioko Sanji, Yuka Kato, Akira Matsuda
Graduate School of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
a r t i c l e i n f o
a b s t r a c t
We describe herein a method for isolating fully-modified 40-thioRNA aptamers against human
a-throm-
Article history:
Received 1 August 2008
Revised 16 September 2008
Accepted 17 September 2008
Available online 20 September 2008
bin using the SELEX protocol. In order to isolate the desired aptamers, in vitro transcription was exam-
ined in the presence of four kinds of 40-thioribonucleoside triphosphates (40-thioNTPs) in an effort to
afford the fully-modified 40-thioRNA. The transcription efficiency of the 40-thioNTPs was first compared
with that of the nucleoside 50-(
the
a-thio)triphosphates (
aSNTPs) and found to be less effective than that of
a
SNTPs, especially when GTP and/or ATP were substituted for 40-thioGTP and/or 40-thioATP. Further
Keywords:
SELEX
Aptamer
40-thioNTP
40-thioRNA
attempts to improve its efficiency, including the use of a mutant RNA polymerase instead of the wild type,
various additives, and 40-thioNTP concentrations were unsuccessful. Accordingly, the transcription was
performed in the presence of 40-thioNTPs together with the natural GTP and ATP at the appropriate con-
centrations. Although this attempt furnished a highly-modified 40-thioRNA, but not a fully-modified 40-
thioRNA, we eventually succeeded in isolating the fully-modified 40-thioRNA aptamers by SELEX using
optimized transcription conditions, followed by post-modification of the resulting aptamers.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
nucleoside triphosphates (NTPs). Thus far, 20-modified-20-deoxy-
nucleoside 50-triphosphates,6 nucleoside 50-(
a
-P-borano)triphos-
SELEX (systematic evolution of ligands by exponential enrich-
ment) is a biotechnological method used for isolating nucleic acid
ligands,1 aptamers, from a random pool of single-stranded nucleic
acid libraries. Since the resulting aptamers bind to the target mol-
ecules, including proteins, with high affinity and specificity in the
same manner as monoclonal antibodies, they are expected to be-
come useful biological tools and diagnostic agents.2,3 In addition,
considering that aptamers often inhibit the function of proteins,
these molecules should become potential therapeutic agents.2,4 Re-
cently, Macugen, an aptamer to vascular endothelial growth factor
(VEGF), was approved as the first therapeutic aptamer.5 For the
above reasons, the use of SELEX to isolate aptamers is an area cur-
rently under intense study.
phates,7 and nucleoside 50-(
a
-thio)triphosphates (
a
SNTPs)8 are
used as modified NTPs for SELEX. Our group has recently reported
the successful isolation of modified RNA aptamers using 40-thiopy-
rimidine nucleoside triphosphates (40-thioUTP and 40-thioCTP;
Fig. 1).9 However, since the substrate susceptibility of NTPs by T7
RNA polymerase is rather restricted, the analogs applicable to SE-
LEX are limited. This being the case, SELEX is generally conducted
using two kinds of modified NTPs, such as UTP and CTP analogs, to-
gether with natural ATP and GTP for effective transcription by the
RNA polymerase. In fact, the aforementioned Macugen was first se-
lected from a modified RNA library transcribed in the presence of
20-FdUTP and 20-FdCTP along with the natural ATP and GTP.10 To
confer higher nuclease stability, most of the natural A and G resi-
In order for aptamers to be suitable for versatile applications,
they must have higher stability in biological fluids and higher bind-
ing affinity. Since natural nucleic acids, especially RNA, is unstable
because of its susceptibility to nuclease digestion, stabilization by
chemical modification of RNA aptamers would be absolutely re-
quired. A highly reliable method of isolating the stabilized RNA
aptamers appears to conduct the SELEX protocol, that is, an
in vitro transcription step in the presence of chemically modified
O
P
O—
O
P
O—
O
P
O—
B
HO
O
O
O
S
HO
OH
4'-thioUTP: B = uracil-1-yl
4'-thioCTP: B = cytosin-1-yl
4'-thioATP: B = adenin-9-yl
4'-thioGTP: B = guanin-9-yl
* Corresponding authors. Tel.: +81 11 706 3228; fax: +81 11 706 4980.
E-mail addresses: noriaki@pharm.hokudai.ac.jp (N. Minakawa), matuda@pharm.
hokudai.ac.jp (A. Matsuda).
Figure 1. Structure of 40-thioNTPs.
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.09.048

N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
9451
dues of the resulting prototype RNA aptamer were post-modified
with 20-OMeA and 20-OMeG nucleotides.5 Consequently, achieve-
ment of SELEX using four kinds of modified NTPs to afford fully-
modified RNA aptamers is considered to be beneficial and
challenging.
tion is +11 and that of the C residue is +13. In fact, a sufficient
amount of the transcript (90% yield) was obtained in the presence
of 40-thioUTP, 40-thioCTP, ATP, and GTP compared with that of the
four natural NTPs.9 Using this template, the in vitro transcription
was examined under various conditions. As shown in Figure 3,
the transcription in the presence of either 40-thioCTP (lane 7) or
As described above, our group has succeeded in isolating par-
tially-modified 40-thioRNA aptamers with potent binding affinity
aSCTP (lane 2) afforded the corresponding transcripts in good
to human
a
-thrombin using 40-thioUTP and 40-thioCTP.9 We also
yield. On the other hand, the transcription using ATP and/or GTP
demonstrated that fully-modified 40-thioRNA prepared by a chem-
ical approach is 600 times more stable than the natural RNA in
human serum.11 Accordingly, we planned to isolate fully-modified
40-thioRNA aptamers by SELEX using four kinds of 40-thioNTPs (40-
thioUTP, 40-thioCTP, 40-thioATP, and 40-thioGTP; Fig. 1). When we
began this research project, only one example employing four
analogs showed a large difference. Thus, contrary to the reactions
using a
SNTP(s) (lanes 3–6), the efficiencies of 40-thioNTP(s) (lanes
8–11) were drastically decreased. This result showed that incorpo-
ration of 40-thioNTPs, especially in the initiation phase, was not tol-
erated for the in vitro transcription by the T7 RNA polymerase,
while that of
oNTPs and SNTPs have a sulfur atom in place of an oxygen atom
a
SNTPs was partially tolerated. Although both 40-thi-
kinds of
a
SNTPs to afford phosphorothiolated RNA aptamers was
a
known.8 Based on these results for isolating chemically modified
RNA aptamers, we carried out an investigation into the selection
of fully-modified 40-thioRNA aptamers.
in their structures, it was revealed that modification of the 40-oxy-
gen to a sulfur atom had a greater effect on the transcription effi-
ciency. The crystal structure of T7 RNA polymerase has been
solved,15 and its interaction with the template and incoming NTP
has been discussed based on the structure of a homologous T7
DNA polymease.16 According to this report, an interaction of the
triphosphate oxygens with the metal (Mg2+) and the RNA polymer-
ase protein (the actual amino acid residues have not been defined)
has been suggested. However, such interactions still seem to be
operative when the oxygen on phosphorus was substituted to a
sulfur atom. On the other hand, little information is reported for
2. Results and discussion
2.1. Screening of reaction conditions to afford 40-thioRNA
transcripts
In order to isolate fully-modified 40-thioRNA aptamers, a 40-thi-
oRNA library has to be generated by in vitro transcription in the
presence of four kinds of 40-thioNTPs. The transcription step by
T7 RNA polymerase comprises two phases,12 the initiation phase
(i.e., starting from +1 to ca. +10 nt in length) and the elongation
phase (i.e., further elongation to a full-length transcript after the
initiation phase). Since the ternary complex consisting of the poly-
merase, the template DNA, and the resulting RNA is unstable in the
initiation phase, formation of abortive transcripts is often observed
in this phase, especially when modified NTPs are incorporated.9,13
Accordingly, successful transcription in the presence of four mod-
ified NTPs appears to be difficult, and no example to afford fully-
modified transcripts has been reported to date, with one exception.
Thus, Ueda et al. examined in vitro transcription by T7 RNA poly-
merase using aSNTPs and showed that fully-modified phosphoro-
thiolated RNAs could be obtained in 33% yield compared with the
reaction in the presence of natural NTPs.14 As a result of the rather
higher transcription efficiency, the isolation of the fully-modified
phosphorothiolated RNA aptamers was successfully conducted by
Jhaveri et al.8
Figure 3. Comparison of transcription efficiency (
transcriptions were carried out using various NTPs concentrations as described in
Section 3; natural NTPs (lane 1); GTP, ATP, UTP plus
SCTP (lane 2) or 40-thioCTP
(lane 7); GTP, CTP, and UTP plus
SATP (lane 3) or 40-thioATP (lane 8); ATP, CTP, and
UTP plus SGTP and SATP
SGTP (lane 4) or 40-thioGTP (lane 9); CTP and UTP plus
SATP, and SCTP
a
SNTPs vs 40-thioNTPs). In vitro
We first examined the transcription in the presence of the 40-
a
thioNTPs, and compared their efficiencies with those of
aSNTPs.
a
In Figure 2, the sequences of the template DNA (template 1) used
in this study and the resulting 30mer RNA (RNA 1) were shown.
Template 1 was designed for effective incorporation of 40-thioUTP
and 40-thioCTP since the initial position of the U residue incorpora-
a
a
a
a
a
(lane 5) or 40-thioGTP and 40-thioATP (lane 10); UTP plus
aSGTP,
(lane 6) or 40-thioGTP, 40-thioATP, and 40-thioCTP (lane 11). The amount of full-
length RNAs was estimated in % based on the conditions in lane 1.
Figure 2. Sequences of DNA templates and the resulting RNAs by in vitro transcription. The italic regions were T7 promoter sequences.

9452
N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
the interaction with the 40-oxygen. From our previous investiga-
tion, the effective incorporation of 40-thioNTPs, that is 40-thioUTP
and 40-thioCTP, was observed in the elongation phase;9 however
this did not take place in the initiation phase. Accordingly, more re-
stricted recognition of the 40-position of the incoming NTPs ap-
pears to occur in this phase.
100
100
90
80
70
60
50
40
30
20
10
0
54
46
In order to improve the transcription efficiency, the following
attempts were examined: (1) utilization of a mutant T7 RNA poly-
merase,17 (2) addition of manganese chloride instead of magne-
sium chloride,18 and (3) conducting the reaction at a higher
spermidine concentration.19 However, none of the above efforts
improved the transcription efficiency (data not shown). As an
alternative, the transcription was conducted at higher modified
NTP concentrations. When the concentration of 40-thioATP was in-
creased to 2.0 mM from the usual 0.4 mM, no alteration of its effi-
ciency was observed (10% vs 11%), while an increase of 40-thioGTP
concentration resulted in the loss of efficiency (10% vs < 1%). Since
it is known that GTP, rather than the other NTP substrates, binds
very tightly to T7 RNA polymerase in the initiation phase,20 a mod-
ified GTP analog, that is 40-thioGTP, at a higher concentration, may
act as an inhibitor of the transcription. From these results, it was
concluded that the in vitro transcription in the presence of four thi-
oNTPs to afford fully-modified 40-thioRNAs would be difficult.
19
8.3
1.1
column 3
column 2
1
2
column 1
3
4
(0:1)
5
(1:9)
6
(1:3)
(1:1)
(3:1)
(1:0)
lane
(NTP:4'-thioNTP)
Figure 4. Comparison of transcription efficiency in the presence of four kinds of 40-
thioNTPs together with natural GTP and/or ATP. In vitro transcription was carried
out under various NTP concentrations as described in Section 3; 40-thioNTPs plus
GTP (column 1; white bars); 40-thioNTPs plus ATP (column 2; black bars); 40-
thioNTPs plus GTP and ATP (column 3; gray bars). The ratio in the parentheses is
GTP:40-thioGTP and/or ATP:40-thioATP. The amount of full-length RNAs was
estimated in % based on the conditions in lane 6 of column 3.
59mer transcript (RNA 2) was prepared (Fig. 2). In vitro transcrip-
tion by T7 RNA polymerase under various NTP concentrations was
examined in a similar manner to those in Fig. 3 (when the tran-
scription goes well, the resulting transcript consists of 26 guanine
nucleotides and 17 adenine nucleotides among the 59 nucleotides).
Then, the resulting full-length transcripts were purified, and
hydrolyzed enzymatically, and then, the composition of the
nucleosides was analyzed by HPLC. These results are summarized
in Figure 5. Thus, the transcription in the presence of 7:1 of
GTP:40-thioGTP and ATP:40-thioATP afforded 72% of the transcript
relative to the conditions in the absence of 40-thioGTP and 40-thi-
oATP. Under these conditions, the incorporation ratios were
estimated as about 6:1 for GTP:40-thioGTP and 16:1 for ATP:40-thi-
oATP from HPLC analysis. As the ratios of 40-thioGTP and 40-thi-
oATP were gradually increased, the incorporation ratios of
40-thioGTP and 40-thioATP also increased, although the transcrip-
tion efficiency decreased (conditions 2 and 3). As can be seen from
these results, the incorporation of 40-thioATP versus the natural
ATP is rather hard to obtain compared to that of 40-thioGTP versus
the natural GTP. In order to improve the ratio of the 40-thioATP
2.2. Transcription in the presence of GTP and ATP
During the course of our investigation, Burmeister et al. suc-
ceeded in isolating fully-modified 20-OMeRNA aptamers using a
mutant T7 RNA polymerase by which NTP analogs modified at
their 20-position are recognized as substrates.21 However, even in
their case, the generation of a fully-modified 20-OMeRNA library
for SELEX was unsuccessful. Accordingly, they carried out in vitro
transcription in the presence of 20-OMeNTPs together with a small
amount of the natural GTP. Under these conditions, it was expected
that the natural GTP would be preferentially incorporated in the
initiation phase, and the 20-OMeNTPs, including 20-OMeGTP, would
be preferentially used for further elongation. As a result, the
authors succeeded in generating a 20-OMeRNA library (this is not
a fully-modified but a highly-modified RNA library), and finally ob-
tained fully-modified 20-OMeRNA aptamers after optimization.21
With their successful results as a reference, we examined in vitro
transcription by T7 RNA polymerase (wild type). Thus, the reaction
was performed in the presence of 40-thioNTPs together with natu-
ral ATP and/or GTP (0.4 mM each as the final NTP concentration),
and the transcription efficiencies at various NTP concentrations
were determined by PAGE analysis. The results are summarized
in Figure 4, in which white and black bars show the results in
the presence of four kinds of 40-thioNTPs together with the natural
GTP (column 1) and ATP (column 2), respectively. The gray bars
(column 3) indicate the results using both GTP and ATP. The num-
bers in the parentheses listed along with the lane numbers are the
ratios of natural NTP:40-thioNTP. As can be seen, no full-length
transcript was observed in lane 1 (NTP:40-thioNTP = 0:1). When
the ratio of NTP:40-thioNTP was changed to 1:9, the transcription
efficiencies were somewhat increased under all conditions (lane
2). However, no obvious improvement was observed in the pres-
ence of either GTP or ATP even if the ratio of NTP was increased
(lanes 3–6 of columns 1 and 2). On the other hand, a drastic
improvement of the transcription efficiency was observed when
both ATP and GTP were included in a ratio of more than 1:1 (lanes
4 and 5 of column 3).
incorporation, the transcription reaction was conducted at higher
40-thioATP concentrations22
(
in Fig. 5, condition 4; 2 mM of 40-
*
thioATP and the ratio of ATP:40-thioATP is 1:3). As a result, the
incorporation ratio was improved to 2:1 for ATP:40-thioATP with-
out a drastic loss in transcription efficiency. Throughout these
experiments, we were able to generate highly-modified 40-thioRNA
transcripts, and thus started an investigation to isolate the fully-
modified 40-thioRNA aptamers.
2.3. SELEX for the selection of fully-modified 40-thioRNA
aptamers against human a-thrombin
As described above, we showed that in vitro transcription in the
presence of four kinds of 40-thioNTPs, together with appropriate
amounts of the natural ATP and GTP, afforded highly-modified
40-thioRNA transcripts. Accordingly, selection of the 40-thioRNA
aptamers under optimized NTP concentrations (condition 4 in
Fig. 5) was next examined. As a target protein for SELEX, human
Since the generation of 40-thioRNA transcripts was successful in
the presence of natural GTP and ATP, we next investigated what ra-
tio of 40-thioGTP and 40-thioATP against GTP and ATP would allow
for competitive incorporation in the resulting transcripts. In order
to check the ratio, a longer template DNA (template 2) giving a
a-thrombin was chosen since we had already isolated its aptamers
using 40-thioUTP and 40-thioCTP.9 Single-stranded DNA (ssDNA)
templates comprising a 30 nt in randomized positions flanked by
constant regions were prepared. After amplification by PCR, the
resulting double-stranded DNA (dsDNA) templates were tran-

N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
9453
Figure 5. HPLC analysis of nucleosides composition of the resulting transcripts under various NTP concentrations. The NTP concentrations in every condition were presented
in Section 3. The amount of full-length RNAs was estimated in % based on the conditions in the presence of GTP and ATP plus 40-thioUTP and 40-thioCTP.
scribed by T7 RNA polymerase in the presence of 40-thioNTPs to-
gether with ATP and GTP to afford a 40-thioRNA library. In general,
the transcription by T7 RNA polymerase afforded sufficient
amounts of the RNA library throughout the selection. However,
in this case, the library obtained at the early stage for the SELEX cy-
cle was too small to subject it to selection even under optimized
NTP concentrations. Therefore, the resulting library was labeled
at the 50-end with 32P, and the resulting radio-labeled library was
used for selections of up to the fifth round of the SELEX cycle. From
the sixth round, the transcription afforded a sufficient amount of
the library, and the selection could be done without radio-labeling
of the library. In all selections, the natural RNA aptamer against the
in which the most abundant sequences represented 15 of the 31
sequences (Fig. 6). These sequences were not identical with our
previous one (i.e., aptamers isolated using 40-thioUTP and 40-thi-
oCTP).9 The predicted sequences, including the constant regions,
were prepared by in vitro transcription under the optimal NTP con-
centrations used for SELEX (i.e., 40-thioNTPs together with GTP and
ATP), and the affinities of the individual full-length transcripts to-
ward human a-thrombin were tested by nitrocellulose filter-bind-
ing assay. All sequences showed binding affinities to the target
protein. Among them, CI-5 from class I, CII-4 from class II, and CIII
were evaluated for their binding affinity. These sequences showed
Kd values of 7.2–16.5 nM (Fig. 7), which were similar to that of the
aptamer isolated using 40-thioUTP and 40-thioCTP.9 As one might
predict, the incorporation of 40-thioGTP (or 40-thioATP) competes
with GTP (or ATP) for in vitro transcription throughout the SELEX
cycles. Therefore, the nucleotide composition of the resulting tran-
scripts, that is the ratio and position of 40-thioGTP (or 40-thioATP)
incorporation, should change in every transcription step if the se-
quence is the same. Despite such bias under our SELEX conditions,
the fact that the library consists of random sequences that con-
verged on the aptamer sequences shows the viability of this SELEX
strategy.
human
a
-thrombin (50-UCCGGAUCGAAGUUAGUAGGCGGA-30)23
was mixed with the generated library to increase a selection bias
and to isolate more potent aptamers. For the selection processes
toward the target protein, a nitrocellulose filter was used, and
the counterselections toward the filter were conducted at every
round of the SELEX cycle to remove the filter-binding species.
The selection processes were repeated under various conditions
(see Section 3 for details), and the library from the eighth round
was reverse transcribed, amplified, and cloned. Accordingly, three
major classes, I, II, and III, emerged from analysis of the 31 clones,
Figure 6. Sequences identified in the eighth round library derived from the affinity selection performed on human a-thrombin. The frequency of clones carrying the same
sequence is indicated in parenthesis. The sequences common to all ligands (constant regions) are shown in lower case letter and the sequences of the random region in upper
case letters.

9454
N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
higher nuclease resistance almost equal to that of 20-OMeRNA
CI-5 (10.5 nM)
CII-4 (7.2 nM)
CIII (16.5 nM)
CI-5-37 (29.6 nM)
90
80
70
60
50
40
30
20
10
0
(unpublished results) and thermally stable duplex formation,11
our results presented here should encourage the development of
nuclease-resistant and thermally stable functional RNA molecules
by SELEX using commercially available and inexpensive wild type
T7 RNA polymerase.
3. Experimental
3.1. General
Physical data were measured as follows: 1H, 13C, and 31P NMR
spectra were recorded on JEOL-EX270 or JEOL-AL400 instruments
in CDCl3, DMSO-d6, or D2O as the solvent with tetramethylsilane
10-10
10-9
10-8
10-7
10-6
(1H and 13C) or H3PO4 31P) as an internal standard. Mass spectra
(
Thrombin (M)
were measured on JEOL JMS-D300 spectrometer. UV spectra were
measure on Shimadzu UV-2450 spectrophotometer. HPLC was per-
formed on Shimadzu CLASS-LC10 unit using a J’sphere ODN 80 col-
umn (4.6 ꢀ 150 mm, YMC). TLC was done on Merck Kieselgel F254
precoated plates. Silica gel used for column chromatography was
Merck silica gel 5715. T7 RNA polymerase and TaKaRa Ex TaqTM
for PCR were purchased from TaKaRa. SupercsriptTM II (RNase Hꢁ)
Reverse Transcriptase and RNaseOUTTM were from Invitrogen.
Nitrocellulose filter (HAWP filter, 0.45 mm) was from Millipore.
Figure 7. The binding of CI-5, CII-4, CIII, and CI-5-37 (see Fig. 8) to varying
concentration of human -thrombin was determined by nitrocellulose filter
partitioning as described in Section 3. The CI-5-37 is a minimized one of CI-5 and
was chemically synthesized as a fully-modified 40-thioRNA.
a
Since the resulting aptamers include the constant regions along
with the random region, minimization of the resulting aptamers is
generally performed in order to optimize their sequences. In our
investigation, it was also necessary to examine whether the mini-
mized sequences still showed potent binding affinity to the target
protein when these were prepared as fully-modified 40-thioRNA
sequences. Therefore, the secondary structure prediction for CI-5,
for example, was made using the Zuker RNA mfold computer algo-
rithm24 to afford a hairpin structure possessing a long stem region.
Although it is not clear whether the predicted structure is the ac-
tual secondary structure or if it is the best length aptamer, we
chemically synthesized a 37mer of fully-modified 40-thioRNA se-
quence (CI-5-37) as a candidate of the minimized sequence based
on the predicted secondary structure (Fig. 8). As a result, the Kd va-
lue of CI-5-37 was estimated as 29.6 nM, although its affinity was
slightly lower than the original CI-5 and its maximal binding activ-
ity was also decreased (Fig. 7). Since post-modification of the RNA
aptamer to the fully-modified 40-thioRNA sequence resulted in
complete loss of its binding affinity in our previous investigation,9
the fact that the CI-5-37 consisting of a fully-modified 40-thioRNA
Human
a-thrombin was from Enzyme Research Laboratories.
pGEM-T Easy Vector System was from Promega. BigDye Termina-
tor v3.1 Cycle Sequence Kit was from Applied Biosystems.
[a- c-
32P]UTP and [ 32P]GTP were purchased from PerkinElmer.
Unlabeled NTPs were from Amersham. Modified NTPs were from
TriLink. Radioactive densities of the gel were visualized by a Bio-
imaging analyzer (Bas 2500, Fuji). DNA was sequenced using a Big-
Dye Terminator v3.1 Cycle Sequence Kit (Applied Biosystems) on a
377 DNA sequencer (Applied Biosystems). DNA templates for tran-
scription were purchased from SIGMA Genosys and RNA aptamer
against a-thrombin was from Hokkaido System Science.
3.2. Synthesis of 40-thioGTP and 40-thioATP
Both compounds were prepared starting from appropriately
protected 40-thionucleoside derivatives according to our previous
method.9,25 The synthetic scheme and experimental details were
presented in the Supplementary material.
showed binding affinity toward human
a-thrombin is worth not-
ing. Although further optimization, including structure and length
of the resulting aptamer, would likely be required to improve its
binding affinity, our results demonstrated that SELEX in the pres-
ence of four kinds of thioNTPs together with appropriate amounts
of natural GTP and ATP is versatile enough to isolate fully-modified
40-thioRNA aptamers after optimization. Since 40-thioRNA showed
3.3. Comparison of transcription efficiency of 40-thioNTPs with
aSNTPs (investigation shown in Fig. 3)
The in vitro transcription was performed with 40 pmol of tem-
plate 1 in 20
0.4 mM of natural NTP(s) and modified NTP(s) including [
(10 Ci), 100 U of T7 RNA polymerase in a 40 mM Tris–HCl buffer
l
L of transcription buffer. The mixture contained either
a
-
32P]UTP
g u
a
a
l
a
a
c
g
a
(pH 8.0) containing 8 mM MgCl2 and 2 mM spermidine, 40 U of
A
G
U
g
g
a
g g
U U U CG U GC C U U U U C U C G
c c a u g g
G
G
RNaseOUTTM, and 5 mM DTT. Reaction mixtures were incubated at
g g g a a g a g u
a
c
a u
a
g
A
37 °C for 3 h, and 2
lL aliquots of the reaction mixture were added
g
a
g
a
to4 L ofloadingbuffer(50 mMEDTA, 10 Murea, 0.1%bromophenol
l
3'
g
g
blue, and 0.1% xylene cyanol). The mixtures were then analyzed by
electrophoresis on 20% polyacrylamide gel containing 7 M urea.
Radioactive densities of the gel were visualized using a Bio-imaging
analyzer, and the transcription efficiencies were estimated in %
based on the conditions in the presence of natural NTPs.
g
CI-5
5'
minimization
A
U
5'
3'
G
G
G
C C U U UU C UC
U C
U
G
G
3.4. In vitro transcription in the presence of 40-thioNTPs
together with GTP and/or ATP (investigation shown in Fig. 4)
g g c a u g g g a a g a g u
a
A
g
G
CI-5-37
The in vitro transcription was performed with 40 pmol of tem-
Figure 8. The predicted secondary structure of CI-5 and its minimized sequence
(CI-5-37). The upper case letter corresponds to the random region and the lower
case letter corresponds to the constant regions.
plate 1 in 20
lL of transcription buffer. The mixture contained
0.4 mM each of 40-thioUTP and 40-thioCTP in every entries, plus

N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
9455
0.4 mM of 40-thioATP and 0.4 mM of [GTP + 40-thioGTP] with
appropriate ratio (column 1), plus 0.4 mM of [ATP + 40-thioATP]
with appropriate ratio and 0.4 mM of GTP (column 2), or plus
0.4 mM each of [GTP + 40-thioGTP] and [ATP + 40-thioATP] with
out 50-end labeling. The resulting library was denatured in anneal-
ing buffer (50 mM Tris–HCl, pH 7.4, 100 mM NaCl, 1 mM MgCl2) at
90 °C for 5 min and allowed to cool at 25 °C for 20 min prior to each
round of selection. In addition, the library was counterselected by
passing through a nitrocellulose filter to remove filter-binding spe-
appropriate ratio (column 3). The mixture also included [
(10 Ci), 100 U of T7 RNA polymerase in a 40 mM Tris–HCl buffer
c-
32P]GTP
l
cies. For the first round of selection, human a-thrombin (1000 nM)
(pH 8.0) containing 8 mM MgCl2 and 2 mM spermidine, 40 U of
was mixed with radio-labeled 40-thioRNA library together with
unlabeled RNA aptamer (50-UCCGGAUCGAAGUUAGUAGGCGGA-
30)23 (5 nM + 500 nM) in the selection buffer (50 mM Tris–HCl,
pH 7.4, 100 mM NaCl, 1 mM MgCl2, 1 mM DTT). After incubation
at 37 °C for 5 min, the mixture was separated by filtration through
a nitrocellulose filter to collect thrombin-binding species and
washed with 1 mL of the selection buffer. The separated species
RNaseOUTTM, and 5 mM DTT. Reaction mixtures were incubated at
37 °C for 3 h, and 2
lL aliquots of the reaction mixture were added
to 4 L of loading buffer (50 mM EDTA, 10 M urea, 0.1% bromophe-
l
nol blue and 0.1% xylene cyanol). The mixtures were then analyzed
by electrophoresis on 20% polyacrylamide gel containing 7 M urea.
Radioactive densities of the gel were visualized using a Bio-imag-
ing analyzer, and the transcription efficiencies were estimated in
% based on the conditions in the presence of 0.4 mM each of 40-thi-
oUTP and 40-thioCTP plus GTP and ATP.
were eluted from the filter by phenol (400
lL) and freshly prepared
7 M urea (200 L). Ethanol precipitation with 20 mg of glycogen
l
was carried out and the 40-thioRNAs recovered were annealed to
reverse primer, and reverse transcribed by SuperscriptTM II (RNase
Hꢁ) at 42 °C for 50 min to give the cDNAs. Following PCR amplifi-
cation, the resulting dsDNA templates were transcribed in vitro
to give the 40-thioRNA library for the next round of selection. The
concentration of the 40-thioRNA library together with unlabeled
RNA aptamer was fixed to approximately 500 nM in each selection,
3.5. Analysis of nucleosides composition of the resulting 59mer
transcripts (investigation shown in Fig. 5)
In the similar manner as described above, in vitro transcription
using template 2 was performed in the presence of 0.4 mM each of
40-thioUTP and 40-thioCTP, plus [GTP + 40-thioGTP] = [0.7 mM +
0.1 mM] and [ATP + 40-thioATP] = [0.7 mM + 0.1 mM] (condition
1), plus [GTP + 40-thioGTP] = [0.6 mM + 0.2 mM] and [ATP + 40-thi-
oATP] = [0.6 mM + 0.2 mM] (condition 2), plus [GTP + 40-thioGTP] =
[0.6 mM + 0.4 mM] and [ATP + 40-thioATP] = [0.6 mM + 0.4 mM]
(condition 3), or plus [GTP + 40-thioGTP] = [0.6 mM + 0.4 mM],
and [ATP + 40-thioATP] = [0.66 mM + 2.0 mM] (condition 4). For
while that of human
a-thrombin in the selection buffer was de-
creased gradually in successive rounds from 1000 nM to 50 nM.
All selection processes were carried out at 37 °C. After eight rounds
of selection, the selected RNA pool was reverse transcribed and PCR
amplified. The resulting DNAs were cloned and sequenced as de-
scribed above. The secondary structure predictions for the selected
aptamers were made using the Zuker RNA mfold computer
algorithm.24
the reaction to estimate the transcription efficiency, [
(10 Ci) was also added to the reaction mixture. After being incu-
c-
32P]GTP
l
bated at 37 °C for 3 h, the transcripts were purified on 20% dena-
turing polyacrylamide gel. The resulting full-length transcripts
(0.2 OD) were incubated with snake venom phosphodiesterase
3.7. Chemical synthesis of thioRNA aptamer
The predicted thioRNA aptamer, CI-5-37, was synthesized on an
Applied Biosystem 3400 DNA synthesizer using four kinds of 40-
thioribonucleoside phosphoramidite units. The synthesis and
purification of the aptamer carried out according to our previous
methods,11 and its structure was confirmed by measurement of
MALDI-TOF/MASS spectrometry on a Voyager-DE pro. CI-5-37: cal-
culated mass, C351H428N143O227P36S37 12583.4 (M–H); observed
mass, 12589.9.
(6 mL), RNase A (10
fer containing 100 mM Tris–HCl (pH 7.7) and 2 mM MgCl2 (total
516 L) at 37 °C for 12 h. After the reaction mixture was heated
lg), and alkaline phosphatase (0.5 U) in a buf-
l
in boiling water for 5 min, the enzymes were removed from reac-
tion mixture by filtration with MicropureÒ-EZ device (Millipore),
and the filtrate was concentrated. Nucleoside composition was
determined by analysis of the residue with reverse-phase HPLC,
using a J’sphere ODN 80 column (4.6 ꢀ 150 mm) with a linear gra-
dient of acetonitrile (from 3.5% to 5% over 30 min) in 0.1 N TEAA
buffer (pH 7.0).
3.8. Filter-binding assay and determination of dissociation
constant
3.6. SELEX protocols
The equilibrium dissociation constants (Kds) for the selected
aptamers, CI-5, CII-4, CIII, and CI-5-37, were determined by using
a constant amount of 50-end labeled oligonucleotides (0.2 nM) in
SELEX was carried out essentially according to reported meth-
ods.9,23 Single-stranded DNA (ssDNA) templates with a 30 nt vari-
able region (50-GGGAGAAGGGAAGTAACAGG-N30-GTGAGAAGAGG
TGACGGTACCAG-30; 73mer), a forward primer (50-GCTCTAGATAA
TACGACTCACTATAGGGAGAAGGGAAGTAACAGG-30; 45mer) and a
reverse primer (50-CTGGTACCGTCACCTCTTCTCAC-30; 23mer) were
prepared. The ssDNA templates were converted to double-
stranded DNA (dsDNA) by PCR with the forward and reverse prim-
ers. The resulting DNA templates were transcribed using T7 RNA
polymerase in the presence of 0.4 mM each of 40-thioUTP and 40-
thioCTP, plus [GTP + 40-thioGTP] = [0.6 mM + 0.4 mM] and [ATP +
40-thioATP] = [0.66 mM + 2.0 mM]. From first to fifth rounds, every
transcriptions were performed using T7 RNA polymerase (250 U)
the selection buffer with increasing concentrations of human
a-
thrombin (0.2–200 nM). After incubation at 37 °C for 5 min, the
mixture was passed through a nitrocellulose filter, and the filter
was immediately washed with selection buffer (200
l
L ꢀ 3).
Radioactivity retained on the filter was quantified by a Bio-imaging
analyzer and evaluated by calculating the percentage of the input
oligonucleotides retained on each nitrocellulose filter in complexes
with protein. The data points were fitted to a Scatchard plot to
determine the equilibrium dissociation constant by PRISM.
Acknowledgments
in a buffer (100
l
L) at 37 °C for 16 h, and the resulting full-length
The authors thank Ms. Y. Misawa and Ms S. Oka (Center for
Instrumental Analysis, Hokkaido University) for technical assis-
tance with the manuscript. This work was supported in part by
Grant-in-Aid for Scientific Research from the Japan Society for Pro-
motion of Science (No. 15510169 to N.M. and Nos. 15209003 and
18109001 to A.M.).
transcripts were purified on 12% denaturing polyacrylamide gel
32
after labeling with P at their 50-end for the selection. From sixth
to eighth rounds, the transcriptions were performed using T7 RNA
polymerase (2500 U) in a buffer (1000
resulting full-length transcripts were used for the selection with-
lL) at 37 °C for 16 h, and the

9456
N. Minakawa et al. / Bioorg. Med. Chem. 16 (2008) 9450–9456
10. Ruckman, J.; Green, L. S.; Beeson, J.; Waugh, S.; Gillette, W. L.; Henninger, D. D.;
Supplementary data
Claesson-Welsh, L.; Janji, N. J. Biol. Chem. 1998, 273, 20556–20567.
11. Hoshika, S.; Minakawa, N.; Matsuda, A. Nucleic Acids Res. 2004, 32, 3815–3825.
12. Yin, Y. W.; Steitz, T. A. Science 2002, 298, 1387–1395.
13. Kujau, M. J.; Siebert, A.; Wolfl, S. J. Biochem. Biophys. Methods 1997, 35, 141–
151.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2008.09.048.
14. Ueda, T.; Tohda, H.; Chikazumi, N.; Eckstein, F.; Watanabe, K. Nucleic Acids Res.
1991, 19, 547–552.
15. Cheetham, G. M. T.; Jeruzalmi, D.; Steitz, T. A. Nature 1999, 399, 80–83.
16. Doublié, S.; Tabor, S.; Long, A. M.; Richardson, C. C.; Ellenberger, T. Nature 1998,
391, 251–258.
17. Brieba, L. G.; Sousa, R. Biochemistry 2000, 39, 919–923.
18. Huang, Y.; Beaudry, A.; McSwiggen, J.; Sousa, R. Biochemistry 1997, 36, 13718–
13728.
References and notes
1. (a) Tuerk, C.; Gold, L. Science 1990, 249, 505–510; (b) Robertson, D. L.; Joyce, G.
F. Nature 1990, 344, 467–468; (c) Ellington, A. D.; Szostak, J. W. Nature 1990,
346, 818–822.
2. Bunka, D. H. J.; Stockley, P. G. Nat. Rev. Microbiol. 2006, 4, 588–596.
3. Jayasena, S. D. Clin. Chem. 1999, 45, 1628–1650.
4. (a) Nimjee, S. M.; Rusconi, C. P.; Sullenger, B. A. Annu. Rev. Med. 2005, 56, 555–
583; (b) Que-Gewirth, N. S.; Sullenger, B. A. Gene Ther. 2007, 14, 283–291.
5. Ng, E. W. M.; Shima, D. T.; Calias, P.; Cunningham, E. T., Jr.; Guyer, D. R.; Adamis,
A. P. Nat. Rev. Drug Discov. 2006, 5, 123–132.
6. Pagratis, N. C.; Bell, C.; Chang, Y.; Jennings, S.; Fitzwater, T.; Jellinek, D.; Dang, C.
Nat. Biotechnol. 1997, 15, 68–73.
7. Lato, S. M.; Ozerova, N. D. S.; He, K.; Sergueeva, Z.; Shaw, B. R.; Burke, D. H.
Nucleic Acids Res. 2002, 30, 1401–1407.
8. Jhaveri, S.; Olwin, B.; Ellington, A. D. Bioorg. Med. Chem. Lett. 1998, 8, 2285–
2290.
9. Kato, Y.; Minakawa, N.; Komatsu, Y.; Kamiya, H.; Ogawa, N.; Harashima, H.;
Matsuda, A. Nucleic Acids Res. 2005, 33, 2942–2951.
19. Padilla, R.; Sousa, R. Nucleic Acids Res. 1999, 27, 1561–1563.
20. Sen, R.; Dasgupta, D. Biochem. Biophys. Res. Commun. 1993, 195, 616–622.
21. Burmeister, P. E.; Lewis, S. D.; Silva, R. F.; Preiss, J. R.; Horwitz, L. R.;
Pendergrast, P. S.; McCauley, T. G.; Kurz, J. C.; Epstein, D. M.; Wilson, C.; Keefe,
A. D. Chem. Biol. 2005, 12, 25–33.
22. As described, in vitro transcription at a higher 40-thioATP concentration does
not affect the efficiency which is not the case at
concentration.
a
higher 40-thioGTP
23. Kubik, M. F.; Stephens, A. W.; Schneider, D.; Marlar, R. A.; Tasset, D. Nucleic
Acids Res. 1994, 22, 2619–2626.
24. Zucker, M. Science 1989, 244, 48–52.
25. Naka, T.; Minakawa, N.; Abe, H.; Kaga, D.; Matsuda, A. J. Am. Chem. Soc. 2000,
125, 7233–7243.