Construction of peptides with nucleobase amino acids: Design and synthesis of the nucleobase-conjugated peptides derived from HIV-1 rev and their binding properties to HIV-1 RRE RNA

Article abstract of DOI:10.1016/S0968-0896(00)00324-2

In order to develop a novel molecule that recognizes a specific structure of RNA, we have attempted to design peptides having L-α-amino acids with a nucleobase at the side chain nucleobase amino acid (NBA), expecting that the function of a nucleobase whic

Full text of DOI:10.1016/S0968-0896(00)00324-2

Bioorganic & Medicinal Chemistry 9 (2001) 991±1000
Construction of Peptides with Nucleobase Amino Acids:
Design and Synthesis of the Nucleobase-Conjugated Peptides
Derived from HIV-1 Rev and their Binding Properties to HIV-1
RRE RNA
Tsuyoshi Takahashi,a Keita Hamasaki,a Akihiko Uenoa and Hisakazu Miharaa,b,
*
aDepartment of Bioengineering, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta,
Midori-Ku, Yokohama 226±8501, Japan
bForm and Function, PRESTO, Japan Science and Technology Corporation, Tokyo Institute of Technology, Nagatsuta,
Midori-Ku, Yokohama 226±8501, Japan
Received 29 August 2000;accepted 21 November 2000
AbstractÐIn order to develop a novel molecule that recognizes a speci®c structure of RNA, we have attempted to design peptides
having l-a-amino acids with a nucleobase at the side chain (nucleobase amino acid (NBA)), expecting that the function of a
nucleobase which can speci®cally recognize a base in RNA is regulated in a peptide conformation. In this study, to demonstrate the
applicability of the NBA units in the peptide to RNA recognition, we designed and synthesized a variety of NBA-conjugated pep-
tides, derived from HIV-1 Rev. Circular dichroism study revealed that the conjugation of the Rev peptide with an NBA unit did not
disturb the peptide conformation. RNA-binding anities of the designed peptides with RRE IIB RNA were dependent on the
structure of the nucleobase moieties in the peptides. The peptide having the cytosine NBA at the position of the Asn40 site in the
Rev showed a higher binding ability for RRE IIB RNA, despite the diminishing the Asn40 function. Furthermore, the peptide
having the guanine NBA at the position of the Arg44 site, which is the most important residue for the RNA binding in the Rev,
bound to RRE IIB RNA in an ability similar to Rev34À50 with native sequence. These results demonstrate that an appropriate NBA
unit in the peptide plays an important role in the RNA binding with a speci®c contact such as hydrogen bonding, and the interac-
tion between the nucleobase in the peptide and the base in the RNA can enhance the RNA-binding anity and speci®city. # 2001
Elsevier Science Ltd. All rights reserved.
Introduction
protein and the nucleobases in an RNA is important to
increase the binding anity and speci®city of the pro-
tein to the RNA.2
RNA±protein interaction plays important roles in nat-
ure. Many cellular processes, including transcription,
RNA splicing, and translation, depend on the speci®c
interaction of proteins and RNA.1 RNAs can fold into
a wide range of tertiary structures, and then RNA-
binding proteins recognize the speci®c structures of
RNA including secondary structural domains such as
hairpin loops, internal loops, and bulges.2 In many
cases, an RNA-binding domain of protein forms a suit-
able conformation such as an a-helix and a b-sheet to
recognize a structured RNA,3À5 and amino acids orien-
tated exactly in the protein structure make speci®c con-
tacts to RNA backbone and bases.2 Especially, the
hydrogen bonding between the amino acid residues in a
On the other hand, interaction between nucleobases is
important to form a speci®c conformation of DNA and
RNA. Since the recognition ability between bases is
highly selective, this ability can be used to make a novel
compound that binds to DNA and RNA speci®cally.
Peptide nucleic acid (PNA), developed by Nielsen and
co-workers, is a DNA mimic with nucleobases on a
pseudopeptide backbone composed of N-(2-amino-
ethyl)glycine units.6,7 A PNA molecule has an ability of
ecient and sequence speci®c binding both single-
stranded DNA and RNA as well as double-stranded
DNA. Although PNA is an useful item to recognize the
sequence of DNA and RNA, a simple PNA molecule
itself may not discriminate a highly structured RNA
with high speci®city, because PNA alone would not
*Corresponding author. Tel.: +81-45-924-5756;fax: +81-45-924-
5833;e-mail: hmihara@bio.titech.ac.jp
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00324-2

992
T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
have a high ability to form various conformations as
proteins or peptides do.
examples of the proteins that bind to RNA speci®cally.
The Rev protein binds to the corresponding responsive
region of HIV-1 mRNA (RRE),16À18 and this protein±
RNA interaction plays a key role in the HIV-1 virus
replication.14,15 The arginine-rich domain (34±50) of the
Rev protein binds speci®cally to the stem-loop IIB
region of RRE RNA (Fig. 1c) by forming an a-helix
conformation.19 NMR structural analyses revealed that
the Rev a-helix penetrates deeply into the major groove
widened by two non-Watoson-Crick base pairs, G47±
A73 and G48-G71.3 The a-helix potential of the
Rev34À50 peptide a€ects the binding anity and speci®-
city of the peptide to RRE IIB RNA.19,20 We have
The study for design and synthesis of molecules that
recognize a speci®c structure of RNA including second-
ary or tertiary structure might lead to a way of drug
discovery targeting RNA. For the goal of developing
such a novel molecule, we have attempted to design
peptides containing l-a-amino acids with a nucleobase
at the side chain (nucleobase amino acid (NBA)),
expecting that the function of a nucleobase capable of
speci®cally recognizing a base in RNA is regulated in a
peptide conformation such as an a-helix. The use of
chiral NBA units should have more advantage than the
use of achiral PNA units in respect of keeping the pep-
tide conformation on RNA-binding.8 We did not select
the l-a-amino-b-(nucleobase)-propanoic acids such as
l-Willardiine,9 which carries a uracil base at the b-posi-
tion of alanine, but l-a-amino-g-(nucleobase)-butyric
acids10,11 as NBA units, expecting the possibility of a
more ¯exible ethylene linker in an e€ective RNA recog-
nition. Furthermore, as an advantage of NBA, it can be
introduced at any position of peptides the same manner
as the natural amino acids by use of the solid-phase
peptide synthesis.
reported that the peptide having cytosine NBA (CNBA
)
instead of Gln36 in Rev34À50 was successfully designed
and synthesized, and the peptide bound to RRE IIB
RNA with higher anity and speci®city than
11
Rev34À50
.
In this study, we designed and synthesized a
series of peptides conjugated with one or two NBA units
at various positions of Rev (Fig. 1), and evaluated the
e€ects of NBA conjugation with the Rev peptide on the
binding anity and speci®city to HIV-1 RRE IIB RNA.
Results and Discussion
To demonstrate the applicability of NBA units in pep-
tides, we have chosen an active part of the regulatory
protein of virion expression (Rev) of human immuno-
de®ciency virus type-1 (HIV-1)12À15 among many
Design and synthesis of the peptides with NBA
Design of a series of NBA-conjugated peptides was
carried out based on the structure of HIV-1 Rev34À50
Figure 1. (a) Amino acid sequences of Rev34À50 and designed NBA-peptides;(b) structures of nucleobase amino acids (NBA), C NBA, TNBA, GNBA
,
and ANBA, and a peptide nucleic acid, CPNA;(c) the secondary structures of wild type RRE IIB RNA, and G48A and C46G74 mutant RNAs.

T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
993
that binds to RRE IIB RNA speci®cally. In many cases
of DNA or RNA binding proteins, Asn and Arg resi-
dues play an important role in making hydrogen bonds
with some nucleic acid bases. It has been proposed by
NMR structural analyses that the amide group of
Asn40 residue in the Rev34À50 interacts with the major
groove edge of both bases of the G47±A73 base pair
(Fig. 1c) through formation of a pair of hydrogen
bonds.3 In order to examine whether the use of a dif-
ferent NBA unit, such as CNBA or TNBA, can modulate
the RNA-binding activity, we designed and synthesized
two peptides, N40CNBA and N40TNBA, in which l-a-
amino-g-cytosine-butyric acid (CNBA) or l-a-amino-g-
thymine-butyric acid (TNBA) was, respectively, intro-
duced in place of Asn40 in Rev34À50 (Fig. 1a). It is
expected that the cytosine or thymine moiety in the
peptide could be substituted for the Asn40 in Rev34À50
owing to the structural similarity. Furthermore, the
chiral CNBA and TNBA units possibly give a rigid con-
formation in the peptides. For the evaluation of struc-
tural importance of NBA, N40CPNA, in which Asn40
and Arg41 were replaced by Nielsen type cytosine PNA
(CPNA),6 was also prepared. The two amino acids were
replaced by CPNA, because a PNA monomer has the
length equivalent to two amino acids in the main chain.
N40A having Ala instead of Asn40 was prepared to
clarify the e€ect of the nucleobase moiety in N40CNBA
and N40TNBA. Moreover, the peptides with two NBA
units at the Gln36 and Asn40 positions were designed
and examined for the ®rst step to construct multi-NBA
peptides. We have already found that the peptide,
Q36CNBA, in which Gln36 is replaced by CNBA, binds to
RRE IIB RNA with a high anity (Kd=1.7 nM).9 On
this basis, Q36CNBAꢀ N40CNBA, in which two CNBA
residues were introduced in places of Gln36 and Asn40,
and Q36CNBAꢀ N40TNBA, in which CNBA and TNBA
were introduced at the 36th and 40th positions, respec-
tively, were designed. It is expected that the second
addition of the CNBA residue at the Gln36 site into the
single mutants, N40CNBA and N40TNBA, is e€ective in
the RNA binding.
group of the Arg side chain and can interact with the
guanine on the G±C pair through hydrogen bonds.
Furthermore, it is also expected that the RNA-binding
speci®city of the peptides with RRE IIB RNA is
enhanced by the conjugation of the GNBA unit due to
diminishing the e€ect of non-speci®c interaction such as
an electrostatic one. To evaluate the e€ect of the NBA-
conjugation, the binding anity of NBA-peptides to
RRE IIB RNA is compared with the Ala-mutant pep-
tides, because the cationic Arg residues have an ability
of the electrostatic interaction with the phosphate
backbone but the NBA units do not have it. To increase
the stability of the a-helix structure, the N-terminal
amino and C-terminal carboxyl groups of the peptides
including Rev34À50 were succinylated and amidated,
respectively.19,20
Fmoc-protected NBA units were synthesized according
to the reported method10 with some modi®cations. The
exocyclic amino group of cytosine and adenine moieties
were protected by the Z group to prevent the side reac-
tion in the peptide synthesis. Peptides were synthesized
by the Fmoc solid-phase method.21 Introduction of the
NBA units did not a€ect the coupling eciency. The
peptides were puri®ed by semi-preparative reversed-
phase HPLC (RP-HPLC) to give a high purity (>98%
on analytical RP-HPLC). The peptides were identi®ed
by matrix assisted laser desorption ionization time-of-
¯ight mass spectrometry (MALDI-TOFMS). The
amino acid analysis was also utilized for determination
of the peptide concentration.
Circular dichroism study of NBA-peptides
Circular dichroism (CD) spectra of a series of the NBA-
.
conjugated peptides were examined in 10 mM Tris HCl
bu€er (pH 7.5) containing 100 mM KCl, 1 mM MgCl2,
and 0.5 mM EDTA or tri¯uoroethanol (TFE) solution
at 25 ꢁC. In the bu€er solution, the conformations of the
peptides, including Rev34À50, were almost in a random
structure (data not shown). It was reported that the Rev
peptide bound to RRE IIB RNA by forming an a-heli-
cal conformation.3,19,20 These results suggest that the
peptides may not form stable hydrogen bonds for
a-helix structure in aqueous solution by the e€ect on the
charge repulsion of the cationic Rev and NBA-peptides,
but the interaction of the peptides with the RNA indu-
ces the a-helical conformation of the peptides. However,
it was dicult to measure the CD spectra of the pep-
tide±RRE IIB RNA complex, because the conformation
of the RNA was also changed by the peptide binding.20
Arg residues in RNA-binding proteins play key roles in
the RNA binding with not only electrostatic interaction
to a phosphate backbone but also hydrogen bonding to
a speci®c base in RNA. The NMR structural analyses
revealed that three out of 10 Arg residues (35th, 39th,
44th positions) in the Rev peptide interacted with the
bases of RRE IIB RNA in a speci®c manner mainly by
making the hydrogen bonds.3 Therefore, the mutation
of the Arg residues to Ala and Lys decreased sig-
ni®cantly the binding anity and speci®city to RRE IIB
RNA.19,20 In order to examine whether the introduction
of various NBA units at the positions of important Arg
residues in the Rev peptide can modulate the RNA-
binding ability, we designed and synthesized the NBA-
conjugated peptides, in which three Arg residues,
Arg35, Arg39, and Arg44 of Rev34À50, were replaced by
the guanine, adenine, cytosine, and thymine NBA units
(Fig. 1a). It is expected that the peptides having an NBA
with guanine (GNBA) bind to the RNA more strongly
than the other peptides, because the guanine moiety
contains functionality comparable to the guanidium
In contrast, in TFE solution, which is known to be an a-
helix forming solvent,22 the NBA-peptides, N40CNBA
and N40TNBA, and Rev34À50 showed a typical a-helix
CD pattern at the amide region (Fig. 2a). The a-heli-
cities of the peptides in TFE were estimated from the
ellipticity at 222 nm as shown in Table 1.23 The a-helix
contents of N40CNBA and N40TNBA were estimated as
63% and 66%, respectively, slightly higher than that of
Rev34À50 (59%). These results suggest that the CNBA
and TNBA units have a potential of forming an a-helix
structure slightly higher than Asn having an a-helix

994
T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
Figure 2. (a) CD spectra of N40CNBA (Ðб), N40TNBA (± ± ± ±), N40CPNA (- - - - - - -), and Rev34À50 (ÐÐÐ) in TFE at 25 ꢁC;(b) CD spectra of
Q36CNBAꢀ N40CNBA (Ðб), Q36CNBAꢀ N40TNBA (- - - -) in TFE at 25 ꢁC. [Peptide]=20 mM.
breaking nature.24 Furthermore, the a-helicities of double
mutants, Q36CNBAꢀ N40CNBA and Q36CNBAꢀ N40TNBA
10 mM Tris±HCl bu€er (pH 7.5) containing 100 mM
KCl, 1 mM MgCl2, and 0.5 mM EDTA at 25 ꢁC. Fluo-
rescence anisotropy of Rhod-Rev was increased by the
addition of RRE IIB RNA. The signi®cant increase of
anisotropy is attributed to the di€erence in molecular
size and conformation between the free and bound
Rhod-Rev with RRE IIB RNA. The dissociation con-
stant of 2.1 nM was calculated from the increasing ani-
sotropy for Rhod-Rev using an equation with 1:1
stoichiometry.
,
were estimated as 54 and 57%, respectively, ca. 10%
lower than N40CNBA and N40TNBA. These results indi-
cate that the introduction of the second NBA unit
slightly decreases the a-helix potential of the peptides.
However, the a-helix potential of the double mutants
was kept as similar level to that of Rev34À50. In contrast,
the a-helicity of N40CPNA was decreased to 30%. We
have reported that the a-helix structure is very much
lowered by the introduction of the ¯exible and achiral
PNA unit in the middle position of the peptides,
although the introduction of a PNA unit at the C-term-
inal of the Rev peptide does not interrupt the a-helix
formation.8 These results indicate that the NBA unit on
the a-helix peptides is superior to the PNA unit in regard
to keeping the peptide conformation.
On the basis of this result, competition experiments
were carried out in the mixture of Rhod-Rev (10 nM)
and RRE IIB RNA (25 nM) by the addition of a pep-
tide solution as a competitor. Fluorescence anisotropy
values were decreased by the addition of Rev34À50 and
the NBA-peptides, a€ording the free Rhod-Rev (Fig. 3).
The dissociation constant was also calculated by an
equation with 1:1 stoichiometry from the anisotropy
decrease. This competition assay revealed that Rev34À50
In the case of Arg-mutant peptides, the NBA-con-
jugated peptides instead of Arg35, Arg39, or Arg44 also
showed a typical a-helix CD pattern in TFE solution.
The a-helicities of the peptides were estimated as about
50±70% (Table 1). The peptides having a cytosine and
thymine moieties generally showed higher a-helicities
than those having guanine and adenine, although the
di€erences were dependent on the NBA positions. These
®ndings indicate that the bulky purine rings on the
peptides inhibit slightly the forming of the a-helical
conformation compared with the pyrimidine rings.
However, the introduction of a NBA unit in the peptide
at each position does not strongly disturb the con-
formational properties of the peptides.
Table 1. a-Helix contents of the peptides in TFE solution
Peptide
Rev34-50
a-Helicity (%)a
Peptide
a-Helicity (%)
59
R35GNBA
R35ANBA
R35CNBA
R35TNBA
R35A
59
55
67
63
65
N40CNBA
N40TNBA
N40A
63
66
59
30
N40CPNA
R39GNBA
R39ANBA
R39CNBA
R39TNBA
R39A
60
54
65
59
55
b
Q36CNBA
Q36TNBA
60
56
54
57
.
Q36CNBA N40CNBA
.
Q36CNBA N40TNBA
The binding activity of NBA-peptides at Asn40 with
RRE IIB RNA
R44GNBA
R44ANBA
R44CNBA
R44TNBA
R44A
54
67
66
67
51
The binding properties of the peptides with RRE IIB
RNA were evaluated by the competition assay25 using
the Rev peptide modi®ed with 5-carboxytetramethyl-
rhodamine at the N-terminal (Rhod-Rev) as a ¯uores-
cence tracer. All the experiments were carried out in
aa-Helix contents were estimated from [y]222 according to the method
of ref 23.
bData from ref 11.

T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
995
Figure 3. (a) Fluorescence anisotropy of Rhod-Rev with RRE IIB RNA as a function of N40CNBA (*), N40TNBA (~), N40CPNA (&), and N40A
(^) concentrations in 10 mM Trisꢀ HCl bu€er (pH 7.5) containing 100 mM KCl, 1 mM MgCl2, and 0.5 mM EDTA at 25 ꢁC. [Rhod-Rev]=10 nM
and [RRE IIB]=25 nM. (b) Relative binding anities between the NBA-peptides and Rev34-50 to RRE IIB RNA. KRd ev/KdNBA is the ratio of the
dissociation constants of Rev34À50 and NBA-peptides.
bound to RRE IIB RNA strongly with a dissociation
constant of 3.4 nM (Table 2). The binding anity of
N40CNBA to the RNA was as strong as that of Rev34À50
(Kd=2.5 nM). N40A bound to the RNA with
Kd=9.4 nM, ca. 4-fold weaker than N40CNBA did.
These results suggest that the cytosine moiety of
N40CNBA contributes to the RNA binding, that is, the
cytosine base in the peptide may interact with the G47±
A73 base pair in the RNA such that the Asn40 residue
in Rev34À50 does.3 In contrast, the binding anity of
N40TNBA to RRE IIB RNA was greatly diminished to
Kd=67 nM, ca. 20-fold weaker than Rev34À50 did, even
though N40TNBA has an a-helix potential similar to
N40CNBA. Although N40A decreased the RNA-binding
anity, ca. 3-fold weaker compared with Rev34À50 by
removing the function of the Asn residue, the thymine
base in N40TNBA signi®cantly weakened the RNA-
binding anity more than Ala in N40A. These ®ndings
indicate that the thymine moiety hardly makes a speci®c
contact such as hydrogen bonding to the RNA, and also
that the existence of the thymine base, which possesses
the bulky and hydrophobic methyl group, in the peptide
further prevents the binding of the peptide with the
RNA. The binding anity of N40CPNA, which has also
the cytosine moiety in the peptide, to RRE IIB RNA
was rather low at Kd=140 nM. This result suggests that
the decrease of the binding anity is attributed to not
only the deletion of the charged residue, Arg41, but also
the lower a-helix potential of N40CPNA (30% in
TFE).8,11,19
potential of Q36CNBAꢀ N40CNBA was lower than that of
N40CNBA. These results indicate that two cytosine
moieties, substituted for Gln36 and Asn40, may interact
with the bases, namely G48 base and G47±A73 base
pair, respectively,3 resulting in keeping the RNA-bind-
ing anity as similar to that of N40CNBA and wild-type
Rev34À50. Q36CNBAꢀ N40TNBA bound to the RNA with
Kd=12 nM, ca. 5-fold stronger than the single mutant
N40TNBA did. This result strongly suggests that the
additional interaction by the cytosine moiety in
.
Q36CNBA N40TNBA e€ectively compensates for the
lower RNA-binding anity of N40TNBA
.
Furthermore, for the elucidation of the binding speci®-
city of Q36CNBAꢀ N40CNBA and Q36CNBAꢀ N40TNBA
for RRE IIB RNA, a mutant G48A RNA, in which the
guanine-48 base of RRE IIB RNA was replaced by
adenine, was also prepared (Fig. 1c). We have reported
that the designed peptide Q36CNBA bound to the mutant
RNA with Kd=5.5 nM (Table 3), 3.2-fold weaker than
that of the wild type RNA (Kd=1.7 nM). In contrast,
Rev34À50 bound to the mutant RNA with Kd=4.6 nM,
Table 2. Dissociation constants (Kd) of the NBA-peptides at Gln36
and Asn40 with wt RRE IIB RNA
a
Peptide
Kd (nM)
KRd ev/KNd BA
Rev34-50
N40CNBA
N40TNBA
N40A
3.4Æ0.3
2.5Æ0.2
67Æ7.6
1.00
1.36
0.05
0.36
0.02
2.00
0.63
1.36
0.28
9.4Æ0.2
140Æ6.2
1.7Æ0.1
5.4Æ0.1
2.5Æ0.3
12.0Æ1.2
N40CPNA
Q36CNBA
Q36TNBA
b
The binding activity of NBA-peptides at Gln36 and
Asn40 with RRE IIB RNA
Q36CNBAꢀ N40CNBA
Q36CNBAꢀ N40TNBA
In the cases of the peptides conjugated with two NBA
.
units, the dissociation constant of Q36CNBA N40CNBA
with RRE IIB RNA was 2.5 nM as same as that of
N40CNBA (Fig. 3b and Table 2), even though the a-helix
aKRd ev/KNd BA is the ratio of dissociation constants of Rev34À50 and
NBA-peptides with wt RRE IIB RNA.
bData from ref 11.

996
T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
comparable to that of the wild type RRE IIB RNA
(Kd=3.4 nM).11 That is, the cytosine moiety in
Q36CNBA may interact with guanine-48 in RRE IIB
RNA, and the interaction enhances the binding speci®-
city to the wild type RRE IIB RNA more than Gln36
in Rev34À50. It is expected that the cytosine moiety
replacing Gln36 in Q36CNBAꢀ N40CNBA also enhances
the binding speci®city to the wild type RNA.
R39ANBA bound to the RNA with Kd=18 nM, ca. 2-fold
stronger than R39A (Table 4). These results suggest that
the introduction of the guanine and adenine at the
position of Arg39 was also e€ective for the RNA bind-
ing as similar to the cases of Arg35 mutants. Interest-
ingly, R39CNBA and R39TNBA showed the dissociation
constants of 81 and 70 nM, ca. 2-fold weaker than
R39A. This result indicates that the introduction of the
cytosine and thymine moieties at the position of Arg39
disturbs the RNA binding. The RNA-binding abilities
of the peptides were signi®cantly dependent on the
structure of the nucleobase moieties in the peptides.
.
Q36CNBA N40CNBA bound to the mutant G48A RNA
with a dissociation constant of 10.0 nM, 4.0-fold weaker
than that of the wild type RNA (Table 3). That is, the
RNA mutation at guanine-48 in RRE IIB RNA
decreases the binding ability of Q36CNBAꢀ N40CNBA
,
but does not change the ability of Rev34À50. The binding
anity of Q36CNBAꢀ N40TNBA to the mutant G48A
RNA was also decreased to Kd=28.0 nM, 2.3-fold
weaker than that of the wild type RNA. These results
indicate that the cytosine moiety at the 36th position
can interact speci®cally with the guanine-48 in RRE IIB
RNA, and the cytosine at the 40th position also has
contacts with the G47±A73 base pair, resulting in that
the binding anity and speci®city was enhanced more
Next, the binding of the Arg44 mutant peptides with
RRE IIB RNA was elucidated. The dissociation con-
stant of R44A with the RNA was 109 nM, ca. 30-fold
weaker than Rev34À50. This result suggests that Arg44 of
the Rev peptide is most important for the binding to
RRE IIB RNA, owing to the signi®cantly speci®c
interaction of the guanidium group on the Arg side
chain with the base in the RNA. Interestingly,
R44GNBA bound to the RNA with Kd=6.9 nM, ca. 16-
fold stronger than R44A did (Fig. 4 and Table 4). This
RNA-binding anity of R44GNBA was almost compar-
able to that of Rev34À50 (2-fold weaker), despite dimin-
ishing the cationic Arg side chain. In contrast, the
binding anities of R44ANBA, R44CNBA, and R44TNBA
were small as Kd=20, 90, and 81 nM, respectively. That
is, only the guanine base is replaceable at the position of
Arg44 so as to maintain the high binding anity. These
results suggest that the guanine moiety in R44GNBA can
make a speci®c contact such as hydrogen bonding with
the RNA.
than that of Rev34À50 and Q36CNBA
.
The binding activity of NBA-peptides at Arg35, Arg39,
and Arg44 with RRE IIB RNA
In the series of Arg35 mutant peptides, R35A having
Ala instead of Arg35 of Rev34À50 bound to RRE IIB
RNA with a dissociation constant of 36 nM, 10-fold
weaker than Rev34À50 did (Kd=3.4 nM). Since Arg35
residue in Rev34À50 interacted not only with the phos-
phate backbone in RRE IIB RNA but also with the
base (guanine-67) in the RNA speci®cally,3 deletion of
the Arg residue decreased signi®cantly the binding a-
nity to the RNA. The NBA-peptides, R35GNBA and
R35ANBA, bound to the RNA with Kd=12 and 9.9 nM,
respectively, ca. 3- or 4-fold stronger than R35A did
(Table 4). In contrast, the binding anities of R35CNBA
and R35TNBA to the RNA were rather low as Kd=29
and 42 nM, a similar anity to R35A. These results
suggest that the guanine and adenine moieties of the
peptides at the 35th position are successful in the RNA
binding, but the cytosine and thymine moieties are not
e€ective for the RNA binding. It seems that the purine
ring can reach the bases in the RNA but the pyrimidine
ring is dicult to interact with the RNA bases.
For the elucidation of the binding speci®city of
R44GNBA to RRE IIB RNA, a mutant RNA C46±G74,
in which G46±C74 base pair was replaced by C46±G74,
was also prepared (Fig. 1c). It was reported by the
NMR structural analyses that the Arg44 side chain in
the Rev peptide interacts with the U45 and is close to
the G46 base in RRE IIB RNA,3 and replacing G46±
C74 with the C46±G74 base pair in RRE IIB RNA
decreased the RNA-binding anity of the Rev pep-
tide.19,20 We have expected that the introduction of the
GNBA unit at the Arg44 site increases the binding speci-
®city to RRE IIB RNA, due to the speci®c interaction of
the guanine moiety in the peptide with the base in RNA.
In the series of Arg39 mutants, R39A bound to RRE
IIB RNA with Kd=38 nM, ca. 10-fold weaker than
Rev34À50 did. R39GNBA bound to the RNA with
Kd=14 nM, ca. 3-fold stronger than R39A did, and
Rev34À50 bound to the mutant C46G74 RNA with a
dissociation constant of 12 nM, 3.5-fold weaker than
Table 4. Dissociation constants (Kd) of the NBA-peptides at Arg35,
Arg39 and Arg44 with RRE IIB RNA
Kd (nM) (KAd la/KNd BA
)
a
Table 3. Dissociation constants (Kd) of the NBA-peptides at Gln36
and Asn40 with the G48A mutant RNA
Peptide
Arg35
Arg39
Arg44
peptide
Kd (nM) for G48A
KGd 48A/Kwd t
a
GNBA
ANBA
CNBA
TNBA
Ala
12Æ1 (3.00)
9.9Æ0.4 (3.63)
29Æ2 (1.24)
42Æ3 (0.86)
36Æ2 (1.00)
14Æ1 (2.71)
18Æ1 (2.11)
81Æ5 (0.47)
70Æ3 (0.54)
38Æ3 (1.00)
6.9Æ0.4 (15.8)
20Æ2 (5.45)
90Æ5 (1.21)
81Æ3 (1.35)
109Æ5 (1.00)
Rev34À50
Q36CNBA
Q36CNBAꢀ N40CNBA
Q36CNBAꢀ N40TNBA
4.5Æ0.3
5.5Æ0.2
1.3
3.2
4.0
2.3
10.0Æ1.0
28.0Æ4.2
aKGd 48A/Kwd t is the ratio of the dissociation constants of peptides with
G48A RNA and wt RRE IIB RNA.
aKAd la/KNd BA is the ratio of the dissociation constants of the peptides
having Ala and NBA-peptides with RRE IIB RNA.

T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
997
Figure 4. (a) Fluorescence anisotropy of Rhod-Rev with RRE IIB RNA as a function of R44GNBA (*), R44CNBA (&), and R44A (~) con-
ꢁ
.
centrations in 10 mM Tris HCl bu€er (pH 7.5) containing 100 mM KCl, 1 mM MgCl2, and 0.5 mM EDTA at 25 C. [Rhod-Rev]=10 nM and [RRE
IIB]=25 nM. (b) Relative binding anities of the Arg44-mutant peptides with RRE IIB RNA. KAd la/KdNBA is the ratio of dissociation constants of
R44A and NBA-peptides.
that to the wild type RRE IIB RNA (Table 5). R44A
also bound to the mutant RNA with Kd=168 nM, 1.5-
fold weaker than that to the wild type RNA. These
results indicate that the interaction of the Arg side chain
with the bases in the RNA enhances the binding speci-
®city to the wild type RNA. Interestingly, R44GNBA
and R44ANBA bound to the mutant RNA with Kd=44
and 122 nM, 6.4- and 6.1-fold, respectively, weaker than
those to the wild type RNA. From the comparison of
the dissociation constants of Rev34À50, R44GNBA, and
R44ANBA it is apparent that the mutation of the G46±
C74 pair to C46±G74 decreases the binding anity of
nine moiety may slightly disturb the structure of the
peptide±RNA complex, resulting in the result that
R44ANBA was not a superior binder to R44GNBA
.
Conclusion
For the aim to construct novel RNA-binding molecules
using the function of nucleobase moieties on the rigid
conformation, we have demonstrated that the Rev-pep-
tides, conjugated with l-a-amino acids with a nucleo-
base, were successfully designed and synthesized, and
bound speci®cally to the RRE IIB RNA. CD studies of
the designed peptides revealed that the NBA-peptides
have a potential to fold in an a-helix structure, although
introduction of the PNA unit at the middle position
inhibits extremely the formation of an a-helix con-
formation. The NBA unit on the a-helix peptide is
superior to the PNA unit in regard to keeping the pep-
tide conformation.
R44GNBA and R44ANBA more than that of Rev34À50
That is, the binding speci®city of R44GNBA and
R44ANBA to the RNA is superior to that of Rev34À50
.
.
On the contrary, binding anities of R44CNBA and
R44TNBA to the mutant RNA were slightly decreased to
Kd=156 and 224 nM, respectively, ca. 2-fold lower than
those to the wild type RNA. Those analyses with the
mutant RNA imply that the guanine moiety of
R44GNBA can make a speci®c interaction with the U45
or G46 base in RRE IIB RNA, and the interaction
enhances the binding speci®city of the peptide with the
RNA. In the case of R44ANBA, the adenine moiety of
R44ANBA can also make a speci®c interaction with the
base in the RNA;however, the introduction of the ade-
The ®rst series of the peptides, in which CNBA and TNBA
,
were introduced instead of Asn40 in the Rev34À50
showed unique RNA-binding properties depending on
the type of the NBA units. The peptide N40CNBA
bound to RRE IIB RNA as a similar level to Rev34À50
despite diminishing the Asn residue. In contrast, the
binding anity of the peptide N40TNBA for the RNA
was lowered much more than N40A. These ®ndings
indicate that not the thymine but the cytosine moiety in
the peptide can interact with the bases in RRE IIB
Table 5. Dissociation constants (Kd) of NBA-peptides with the
C46G74 mutant RNA
a
Peptide
Kd (nM)
KCd 46G74/Kwd t
RNA in a speci®c manner such as Asn in Rev34À50
.
Furthermore, the peptides having two cytosine NBA
units at the positions of Gln36 and Asn40 also showed
the RNA-binding anity similar to native Rev34À50 and
a single mutant N40CNBA. Both cytosine moieties in the
peptide were successful in interacting with the bases in
RRE IIB RNA for keeping the binding activity and
enhancing the binding speci®city. If the cytosine moiety
of the peptides may compensate for the Asn function,3
Rev34-50
12Æ2
44Æ4
3.5
6.4
6.1
1.7
2.3
1.5
R44GNBA
R44ANBA
R44CNBA
R44TNBA
R44A
122Æ10
156Æ18
224Æ17
168Æ29
aKCd 46G74/Kwd t is the ratio of the dissociation constants of peptides with
C46G74 RNA and RRE IIB RNA.

998
T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
exocyclic NH2 proton and N3 atom of cytosine make
hydrogen bonds with O6 oxygen of G47 and exocyclic
NH2 proton of A73, respectively, in the RNA.
with a 100W lamp for 2 h at 20 ꢁC and puri®cation by
column chromatography on silica (80% yield). Boc-
NH-CH(CH2CH2Br)±COOt-Bu (4.6 mmol) was reacted
with N4-benzyloxycarbonylcytosine26 (6.4 mmol) in the
presence of K2CO3, Cs2CO3, and tetrabꢁutylammonium
iodide (TBAI;0.58 mmol) in DMF at 50 C. Puri®cation
by silica gel chromatography gave tert-butyl 4-(N4-
benzyloxycarbonylcytosin-1-yl)-2S-butyloxycarbonyl-
aminobutyrate (Boc-CNBA(Z)±Ot-Bu) (50% yield).
Deprotection of the Boc and t-Bu groups in Boc-
CNBA(Z)±Ot-Bu was carried out with TFA in CH2Cl2
solution at 0 ꢁC, and protection of a-amino group was
reacted with 9-¯uorenylmethyl succinimidyl carbonate
(Fmoc-OSu) to give ®nal material 4-(N4-benzyloxy-
carbonylcytosin-1-yl)-2S-(¯uoren-9-ylmethoxycarbonyl)-
aminobutyric acid (Fmoc-CNBA(Z)±OH) (80% yield).
Reaction of Boc-NH-CH(CH2CH2Br)±COOt-Bu with
N3-benzoylthymine27 in the presence of K2CO3,
Cs2CO3, and TBAI in DMF at 50 ꢁC gave tert-butyl 4-(N3-
benzoylthymin-1-yl)-2S-butyloxycarbonylaminobutyrate
(Boc-TNBA(Bz)±Ot-Bu) (90% yield). Deprotection of
the benzoyl, Boc, and t-Bu groups in Boc-TNBA(Bz)±Ot-
Bu was carried out with 30% HBr in acetic acid at 0 ꢁC,
and protection of a-amino group was reacted with
Fmoc-OSu to give 4-(thymin-1-yl)-2S-(¯uoren-9-ylme-
thoxycarbonyl)aminobutyric acid (Fmoc-TNBA-OH)
(63% yield). Reaction of Boc-NH-CH(CH2CH2Br)-
COOt-Bu with 2-amino-6-chloropurine in the presence
of K2CO3, Cs2CO3, and TBAI in DMF at 50 ꢁC gave
tert-butyl 4-(2-amino-6-chloropurin-9-yl)-2S-butylox-
ycarbonylaminobutyrate (Boc-GNBA(Cl)-Ot-Bu) (62%
yield). Deprotection of the Boc and t-Bu groups
and hydrolysis at position 6 of the purine ring in
Boc-GNBA(Cl)-Ot-Bu were carried out with 1 N HCl
solution for 2 h at 90 ꢁC. The protection of a-amino
group was reacted with Fmoc-OSu to give 4-(guanin-9-
yl)-2S-(¯uoren-9-ylmethoxycarbonyl)aminobutyric acid
(Fmoc-GNBA-OH) (76% yield). Reaction of Boc-NH-
CH(CH2CH2Br)-COOt-Bu with N6-N6-benzyloxycar-
bonyladenine26 in the presence of K2CO3, Cs2CO3, and
TBAI in DMF at 50 ꢁC gave tert-butyl 4-(N6-benzyloxy-
carbonyladenin-9-yl)-2S-butyloxycarbonylaminobutyrate
(Boc-ANBA(Z)-Ot-Bu) (47% yield). After treatment with
TFA and protection by the Fmoc group, 4-(N6-benzyl-
oxycarbonyladenin - 9 - yl) - 2S - (¯uoren - 9 - yl - methoxy-
carbonyl) aminobutyric acid (Fmoc-ANBA(Z)-OH) was
obtained (37% yield). The guanine ring was not pro-
tected, since the amino group on the guanine was hardly
modi®ed under the reaction of elongation of peptides.
These amino acid derivatives were identi®ed by 1H
NMR spectra; 1H NMR (500 MHz, (D6)DMSO, 25 ꢁC):
Fmoc-CNBA(Z)-OH 10.76 (br, 1H), 7.92±7.86 (m, 3H),
7.72 (m, 2H), 7.43±7.32 (m, 10H), 6.98 (d, 1H), 5.18 (s,
2H), 4.32 (m, 2H), 4.24 (m, 1H), 3.93 (m, 1H), 3.83 (m,
2H), 2.15 (m, 1H), 1.94 (m, 1H);Fmoc-T NBA-OH 11.23
(s, 1H), 7.90 (d, 2H), 7.73 (d, 2H), 7.46±7.30 (m, 6H),
4.31 (m, 2H), 4.24 (m, 1H), 3.95 (m, 1H), 3.69 (m, 2H),
The second series of the peptides, in which CNBA, TNBA
,
GNBA, and ANBA, were introduced instead of the Arg35,
Arg39, or Arg44 of the Rev34À50, also showed the RNA-
binding properties depending on the NBA type. In the
Arg35- and Arg39-mutant peptides, the binding ani-
ties of the peptides having the GNBA and ANBA for RRE
IIB RNA were higher than those having CNBA, TNBA
and Ala. Interestingly, in the Arg44-mutant peptides,
only the GNBA mutant, R44GNBA, bound to the RNA
comparable to Rev34À50, although the binding anities
of the other peptides were signi®cantly decreased. The
Arg44 residue of the Rev peptide is the most important
residue for the binding with RRE IIB RNA. Moreover,
only GNBA can replace the Arg function in Rev34À50
,
probably due to its speci®c contacts such as hydrogen
bonding with the bases in the RNA.3
In summary, the NBA units in the a-helix-forming pep-
tide, derived from HIV-1 Rev, can be e€ectively utilized
for the speci®c binding with RRE IIB RNA. The func-
tion of the NBA units orientated on peptide structures
will be further demonstrated for the binding to other
RNAs. This study can lead to a new strategy applicable
to the construction of molecules that speci®cally recog-
nize structured RNAs using the various NBA units on
peptide 2D/3D structures.
Experimental
Materials and methods
All chemicals and solvents were of reagent or HPLC
grade. Amino acid derivatives and reagents for peptide
synthesis were purchased from Watanabe Chemical.
MALDI-TOFMS was measured on
a Shimadzu
MALDI III mass spectrometer by using 3,5-dimethoxy-
4-hydroxycinnamic acid as a matrix. HPLC was carried
out on a YMC ODS A-302 5C18 column (YMC)
(4.6Â150 mm) or a YMC ODS A-323 5C18 column
(10Â250 mm) by employing a Hitachi L-7000 HPLC
system. Amino acid analyses were performed by using
the phenyl isothiocyanate (PTC) method on a Wakopak
WS-PTC column (Wako chemical).
Synthesis of Fmoc-protected ꢀ-amino-ꢁ-(Z-cytosine,
thymine, guanine, and Z-adenine)-butyric acids (Fmoc-
CNBA(Z)-OH, Fmoc-TNBA-OH, Fmoc-GNBA-OH, and
Fmoc-ANBA(Z)-OH)
Fmoc-protected-g-nucleobase amino acids were synthe-
sized according to the reported method10 with some
modi®cations. tert-Butyl 4-bromo-2S-butoxycarbonyl-
aminobutyrate (Boc-NH-CH(CH2CH2Br)-COOt-Bu)
was obtained by the reaction (3 h at 0 ꢁC) of the Boc-
Glu-Ot-Bu (26 mmol) and 2-mercaptopyridine N-oxide
(27 mmol) in the presence of DCC (27 mmol) and di-
methylaminopyridine (2.6 mmol) in a mixture of THF
(10 mL) and CBrCl3 (10 mL) followed by irradiation
2.18 (m, 1H), 1.88 (m, 1H), 1.72 (s, 3H);Fmoc-G
-
NBA
OH 10.58 (s, 1H), 7.90 (d, 2H), 7.78 (d, 1H), 7.72 (d,
2H), 7.62 (s, 1H), 7.43±7.30 (m, 5H), 6.43 (br, 2H), 4.33
(m, 2H), 4.25 (m, 1H), 4.20±3.94 (m, 2H), 3.86 (m, 1H),
2.25 (m, 1H), 2.00 (m, 1H);Fmoc-A NBA(Z)±OH 10.61
(s, 1H), 8.57 (s, 1H), 8.36 (s, 1H), 7.88 (d, 2H), 7.73 (d,

T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
999
2H), 7.46±7.30 (m, 10H), 5.38 (s, 2H), 4.68±4.40 (m,
5H), 3.87 (m, 1H), 2.40 (m, 1H), 2.14 (m, 1H).
cia Biotech). The two PCR primers used were 50-AATT-
TAATACGACTCACTATA-30 (primer 1;T7 promoter
sequence, 21-mer) and 50-GGCTGGCCTGTACCGTC-
AGCG-30 (primer 2, 21-mer). PCR was carried out at
95 ꢁC for 30 s, 55 ꢁC for 30 s and 72 ꢁC for 30 s, and the
product was puri®ed by ethanol precipitation. RRE IIB
RNA was prepared by transcription30 of the DNA
using the RiboMAXTM large scale RNA production
system-T7 (Promega). The template DNA was degraded
by RNase-free DNase I, and then RNA was puri®ed by
Sephadex G-50 gel ®ltration and by polyacrylamide gel
electrophoresis. Other mutant RRE IIB RNAs, G48A
and C46G74 were also prepared by the same method
described above. Concentrations of RNA were deter-
mined by the absorbance at 260 nm.
Peptide synthesis
The peptides were synthesized by the solid-phase
method according to the procedure using 2-(1H-benzo-
triazol-1-yl)-1,1,3,3-tetramethyluronium hexa¯uorophos-
phate (HBTU) and 1-hydroxybenzotriazole hydrate
(HOBtꢀ H2O) as a coupling reagent.21 Fmoc derivatives
of the following amino acids were used: Arg(Mtr),
Asn(Trt), Gln(Trt), Glu(Ot-Bu), and Thr(t-Bu) (Mtr;4-
methoxy-2,3,6-trimethylbenzenesulphonyl, Trt;triphe-
nylmethyl). To synthesize succinylated-peptide resin, the
Fmoc-deprotected peptide resin was treated with succi-
nic anhydride (10 equivÂ2) in 2-methylpyrolidone
(NMP) for 30 min. To remove the resin and protecting
groups, the peptide resin was stirred with 1 M tri-
methylsilyl bromide (TMSBr;0.70 mL) in TFA (4.0 mL)
in the presence of thioanisole (0.36 mL), m-cresol
(0.06 mL), and ethanedithiol (0.18 mL) as a scavenger
for 1.5 h at 0 ꢁC.28 The resin was ®ltered and washed
three times with TFA. The solvent was removed by the
¯ow of N2 gas and solidi®ed with diethyl ether on ice-
bath to give crude peptides. The Z groups on the ade-
nine and the cytosine moieties in the peptides were
stable during the treatment with 1 M TMSBr. The Z
groups were removed with 1 M trimethylsilyl tri-
¯uoromethanesulphonate (TMSOTf;0.93 mL) in TFA
(3.0 mL) in the presence of thioanisole (0.57 mL),
m-cresol (0.10 mL), and ethanedithiol (0.20 mL) for
1.5 h at 0 ꢁC.29 The product was solidi®ed with diethyl
ether on ice-bath. All crude peptides were puri®ed with
CD measurements
CD spectra were recorded on a Jasco J-720WI spectro-
polarimeter using a quartz cell with 1.0 mm pathlength
in the amide region (190±250 nm). Peptides were dis-
.
solved in 10 mM Tris HCl bu€er (pH 7.5) containing
100 mM KCl, 1 mM MgCl2, and 0.5 mM EDTA or tri-
¯uoroethanol (TFE) solution in peptide concentration
of 20 mM.
Fluorescence anisotropy measurements
Fluorescence anisotropy measurements were performed
on a Shimadzu RF-5300PC spectro¯uorophotometer
using a quartz cell with 1.0 cm pathlength at 25 ꢁC in
.
10 mM Tris HCl bu€er (pH 7.5) containing 100 mM
KCl, 1 mM MgCl2, and 0.5 mM EDTA. Fluorescence
anisotropy was calculated by the intensities at 580 nm
excited at 540 nm.
RP-HPLC on
a
YMC ODS A-323 column
(10Â250 mm) using a linear gradient of 5±35% acetoni-
trile/0.1% TFA (30 min) to give puri®ed peptides. Pep-
tides were identi®ed by molecular ion peak (M+H)+ on
MALDI-TOFMS;N40C NBA, m/z 2617.2 ((M+H)+)
Determination of dissociation constants between peptides
and RRE IIB RNA
(calc.=2617.9);N40T
(calc.=2633.0);N40C
(calc.=2518.8);Q36C NBA N40CNBA
((M+H)+) (calc.=2684.0);Q36C NBA N40TNBA, m/z
2698.4 ((M+H)+) (calc.=2699.0);R35G NBA, m/z 2615.9
,
,
m/z 2632.0 ((M+H)+)
m/z 2518.0 ((M+H)+)
NBA
The anities of the peptides to RRE IIB RNA were
determined by the competition assay using ¯uorescence
anisotropy change of Rhod-Rev as a tracer. Mixture of
Rhod-Rev (10 nM) and RRE IIB RNA (25 nM) in the
bu€er solution was titrated with peptide. After each
addition of peptide, samples were stirred for 1 min and
equilibrated for 5 min at 25ꢁC, then ¯uorescence aniso-
tropy was measured. The anisotropy change with increas-
ing peptide concentrations was corrected for dilution.
The dissociation constants of the peptides with RNA
were calculated by eq (1) assumed as a 1:1 stoichiometry
using Kaleida Graph (Synergy Software);
PNA
.
,
m/z 2685.1
.
((M+H)+) (calc.=2615.9);R35A
((M+H)+) (calc.=2599.9);R35C
((M+H)+) (calc.=2575.9);R35T
((M+H)+) (calc.=2590.9);R39G
((M+H)+) (calc.=2615.9);R39A
((M+H)+) (calc.=2599.9);R39C
((M+H)+) (calc.=2575.9);R39T
((M+H)+) (calc.=2590.9);R44G
((M+H)+) (calc.=2615.9);R44A
((M+H)+) (calc.=2599.9);R44C
((M+H)+) (calc.=2575.9);R44T
((M+H)+) (calc.=2590.9).
,
,
,
m/z 2599.1
m/z 2575.9
m/z 2591.4
m/z 2617.0
m/z 2600.3
m/z 2575.5
m/z 2591.4
m/z 2616.4
m/z 2599.3
m/z 2575.9
m/z 2590.7
NBA
NBA
NBA
,
NBA
,
,
,
NBA
NBA
NBA
,
NBA
À
Á
,
,
,
ꢀP† ˆ K0 =ꢀKdꢀA À Af †=ꢀAb À A†† ‡ 1
NBA
0
d
À
NBA
 ꢀR†0 À ꢀKdꢀA À Af †=ꢀAb À A††
À ꢀT†0ꢀA À Af †=ꢀAb À Af †
NBA
Á
ꢀ1†
where (P)0, (R)0, and (T)0 represent the initial con-
centrations of the peptide, RNA, and Rhod-Rev as a
tracer, respectively, and A, Ab, Af are the anisotropy
values of each solution, the bound Rhod-Rev, and the
free Rhod-Rev, respectively. K0d and Kd are the dis-
sociation constants of Rhod-Rev and the peptide with
RNA, respectively.
Preparation of RRE IIB and the mutant RNA
Double-stranded DNA corresponding to RRE IIB
RNA was prepared by the PCR strategy using 50-
AATTTAATACGACTCACTATAGGCTGGTATGG-
GCGCAGCGTCAATGACGCTGACGGTACAGGC-
CAGCC-30 synthetic 68-mer ssDNA (Amersham Pharma-

1000
T. Takahashi et al. / Bioorg. Med. Chem. 9 (2001) 991±1000
16. Bartel, D. P.;Zapp, M. L.;Green, M. R.;Szostak, J. W.
Cell 1991, 67, 529.
References
1. Shiomi, H.;Dreyfuss, G. Curr. Opin. Genet. Dev. 1997, 7,
345.
2. Draper, D. E. J. Mol. Biol. 1999, 293, 255.
3. Battiste, J. L.;Mao, H.;Rao, N. S.;Tan, R.;Muhandiram,
D. R.;Kay, L. E.;Frankel, A. D.;Williamson, J. R. Science
1996, 273, 1547.
17. Cook, K. S.;Fisk, G. J.;Hauber, J.;Usman, N.;Daly, T.
J.;Rusche, J. R. Nucleic Acids Res. 1991, 19, 1577.
18. Kjems, J.;Calnan, B. J.;Frankel, A. D.;Sharp, P. A.
EMBO J. 1992, 11, 1119.
19. Tan, R.;Chen, L.;Buettner, J. A.;Hudson, D.;Frankel,
A. D. Cell 1993, 73, 1031.
4. Ye, X.;Gorin, A.;Ellington, A. D.;Patel, D. J.
Struct. Biol. 1996, 3, 1026.
5. Puglisi, J. D.;Chen, L.;Blanchard, S.;Frankel, A. D.
Science 1995, 270, 1200.
6. Nielsen, P. E.;Haaima, G. Chem. Soc. Rev. 1997, 73.
7. Nielsen, P. E.;Egholm, M.;Berg, R. H.;Buchardt, O.
Science 1991, 254, 1497.
8. (a) Kumagai, I.;Takahashi, T.;Hamasaki, K.;Ueno, A.;
Mihara, H. Bioorg. Med. Chem. Lett. 2000, 10, 377. (b)
Takahashi, T.;Kumagai, I.;Hamasaki, K.;Ueno, A.;Mihara,
H. Nucleic Acids Symp. Ser. 1999, 42, 271.
Nature,
20. Tan, R.;Frankel, A. D. Biochemistry 1994, 33, 14579.
21. Chan, W. C.;White, P. D. In Fmoc Solid Phase Peptide
Synthesis: A Practical Approach;Chan, W. C., White, P. D.,
Eds.;Oxford University Press: New York, 2000;p 41.
22. Zhou, N. E.;Zhu, B.-Y.;Kay, C. M.;Hodges, R. S. Bio-
polymers 1992, 32, 419.
23. Scholtz, J. M.;Qian, H.;York, E. J.;Stewart, J. M.;
Baldwin, R. L. Biopolymers 1991, 31, 1463.
24. Stellwagen, E.;Park, S.-H.;Shalongo, W.;Jain, A.
polymers 1992, 32, 1193.
Bio-
25. (a) Hamasaki, K.;Rando, R. R. Anal. Biochem. 1998, 261,
183. (b) Matsumoto, C.;Hamasaki, K.;Mihara, H.;Ueno, A.
Bioorg. Med. Chem. Lett. 2000, 10, 1857.
9. Lembein, F.;Van Parijs, R. Biochem. Biophys. Res. Com-
mun. 1968, 32, 474.
10. (a) Lenzi, A.;Reginato, G.;Taddei, M. Tetrahedron Lett.
1995, 36, 1713. (b) Ciapetti, P.;Soccolini, F.;Taddei, M. Tetra-
hedron 1997, 53, 1167.
11. Takahashi, T.;Hamasaki, K.;Kumagai, I.;Ueno, A.;
Mihara, H. Chem. Commun. 2000, 349.
26. Thomson, S. A.;Josey, J. A.;Cadilla, R.;Gaul, M. D.;
Hassman, C. F.;Luzzio, M. J.;Pipe, A. J.;Reed, K. L.;Ricca,
K. L.;Wiethe, R. W.;Noble, S. A. Tetrahedron 1995, 51, 6179.
27. Cruickshank, K. A.;Jiricny, J.;Reese, C. B. Tetrahedron
Lett. 1984, 25, 681.
12. Jain, C.;Belasco, J. G. Cell 1996, 87, 115.
13. Malim, M. H.;Hauber, J.;Le, S.-Y.;Maizel, J. V.;Cullen,
B. R. Nature 1989, 338, 254.
14. Felber, B. K.;Hadzopoulou-Cladaras, M.;Cladaras, C.;
Copeland, T.;Pavlakis, G. N. Proc. Natl. Acad. Sci. U.S.A.
1989, 86, 1495.
28. Funakoshi, S.;Murayama, E.;Guo, L.;Fujii, N.;Yajima,
H. J. Chem Soc., Chem. Commun. 1988, 382.
29. Fujii, N.;Otaka, A.;Ikemura, O.;Akaji, K.;Funakoshi,
S.;Hayashi, Y.;Kuroda, Y.;Yajima, H. J. Chem. Soc., Chem.
Commun. 1987, 274.
30. Milligan, J. F.;Uhlenbeck, O. C. Methods Enzymol. 1989,
180, 51.
15. Emerman, M.;Vazeux, R.;Peden, K. Cell 1989, 57, 1155.