Molecular Design for a Pinpoint RNA Scission. Interposition of Oligoamines between Two DNA Oligomers

Article abstract of DOI:10.1021/jo9611780

Phosphoramidite monomers having a benzyl or ethyl ester in the side chain have been synthesized for the facile and selective functionalization of oligonucleotides. Various amino compounds were incorporated into the desired sites in oligonucleotides by ami

Full text of DOI:10.1021/jo9611780

846
J . Org. Chem. 1997, 62, 846-852
Molecu la r Design for a P in p oin t RNA Scission . In ter p osition of
Oligoa m in es betw een Tw o DNA Oligom er s1
Masayuki Endo, Yasushi Azuma, Yoshitaka Saga, Akinori Kuzuya, Gota Kawai, and
Makoto Komiyama*
Department of Chemistry and Biotechnology, Graduate School of Engineering, The University of Tokyo,
Hongo, Bunkyo-ku, Tokyo 113, J apan
Received J une 24, 1996X
Phosphoramidite monomers having a benzyl or ethyl ester in the side chain have been synthesized
for the facile and selective functionalization of oligonucleotides. Various amino compounds were
incorporated into the desired sites in oligonucleotides by aminolysis of the esters. The conjugates,
in which oligoamines (catalysts for RNA hydrolysis) were interposed between two oligonucleotides,
hydrolyzed RNA substrates at exactly the target sites, which was consistent with the results of a
computer-modeling study.
In tr od u ction
Recently, artificial ribonucleases have been attracting
attention, owing to their potential use as tools for gene
manipulation, the control of gene expression, and others.2-5
In most cases, RNA-cleaving molecules are bound to the
terminals of oligonucleotides, which bind to a specific
sequence in the target RNA. The scission by these
conjugates takes place at a limited number of sites.4
However, a precise prediction of the scission site is
difficult, due to the dangling motion of the single-
stranded portion of the substrate RNA. In addition, a
wobbling of the cleaver tends to render the scission less
selective.
In order to cut RNA still more selectively at a desired
site, it is preferable to attach cleaving molecules inside
of the DNA strands. When two portions of the DNA
in these artificial enzymes form heteroduplexes with
the complementary parts in the substrate RNA, the
cleavers are precisely placed at the targeted position.
The molecular movements of the RNA substrates
X Abstract published in Advance ACS Abstracts, J anuary 15, 1997.
(1) A preliminary communication on the synthesis of the phosphora-
midite 3: Endo, M.; Saga, Y.; Komiyama, M. Tetrahedron Lett. 1994,
35, 5879.
(2) (a) Cech, T. R. Science 1987, 236, 1532. (b) Komiyama, M. J .
Biochem. 1995, 118, 665. (c) DeMesmaeker, A.; Ha¨ner, R.; Martin, P.;
Moser, H. E. Acc. Chem. Res. 1995, 28, 366.
Figu r e 1. Modified oligonucleotides synthesized in the present
study.
(3) (a) Matthews, J . A.; Kricka, L. J . Anal. Biochem. 1988, 169, 1.
(b) Marcus-Sekura, C. J . Anal. Biochem. 1988, 172, 289. (c) Toulme`,
J .-J ; He´le`ne, C. Gene 1988, 72, 51. (d) Maher, L. J ., III, Wold, B.;
Dervan, P. B. Science 1989, 245, 725.
and artificial enzymes are suppressed in rigidly struc-
tured duplexes, thus making the scission site more
predictable.
(4) (a) Zuckermann, R. N.; Schultz, P. G. J . Am. Chem. Soc. 1988,
110, 6592. (b) Matsumura, K.; Endo, M.; Komiyama, M. J . Chem. Soc.,
Chem. Commun. 1994, 2019. (c) Bashkin, J . K.; Frolova, E. I.; Sampath,
U. J . Am. Chem. Soc. 1994, 116, 5981. (d) Magda, D.; Miller, R. A.;
Sessler, J . L.; Iverson, B. L. J . Am. Chem. Soc. 1994, 116, 7439. (e)
Hall, J .; Hu¨sken, D.; Pieles, U.; Moser, H. E.; Ha¨ner, R. Chem. Biol.
1994, 1, 185. (f) Uchiyama, Y.; Inoue, H.; Ohtsuka, E.; Nakai, C.;
Kanaya, S.; Ueno, Y.; Ikehara, M. Bioconjugate Chem. 1994, 5, 327.
(g) Komiyama, M.; Inokawa, T.; Yoshinari, K. J . Chem. Soc., Chem.
Commun. 1995, 77.
(5) For DNA scission by artificial nucleases, see: (a) Dervan, P. B.
Science 1986, 232, 464. (b) Moser, H. E.; Dervan, P. B. Science 1987,
238, 645. (c) Sigman, D. S.; Chen, C. B. Annu. Rev. Biochem. 1990,
59, 207. (d) Nielsen, P. E. Bioconjugate Chem. 1991, 2, 1. (e) Sigman,
D. S.; Bruice, T. W.; Mazumder, A.; Sutton, C. L. Acc. Chem. Res. 1993,
26, 98. (f) Thuong, N. T.; He´le`ne, C. Angew. Chem., Int. Ed. Engl. 1993,
32, 666. (g) Reference 2b.
In the present study, artificial ribonucleases were
prepared by interposing oligoamines (catalysts for RNA
hydrolysis)8 between two DNA strands (see Figure 1).
Novel phosphoramidite monomers (3 and 6 in Scheme
1) having ester residues in the side chains were synthe-
sized for this purpose. The riboses in the conventional
phosphoramidite monomers were replaced by trimethyl-
ene main chains in order to minimize the steric perturba-
tion. Oligoamines were introduced by post-reactions to
the desired sites in the oligonucleotides using aminolysis
of the esters. These artificial ribonucleases hydrolyzed
substrate RNA at exactly the predicted site. The scission
patterns are discussed in terms of the results of a
computer-modeling study.
(6) (a) Uhlmann, E.; Peyman, A. Chem. Rev. 1990, 90, 543. (b)
Goodchild, J . Bioconjugate Chem. 1990, 1, 165.
(7) Diastereoselective synthesis of a phosphoramidite monomer of
dinucleotide, which has a reactive residue on the phosphorus atom,
was reported: Endo, M.; Komiyama, M. J . Org. Chem. 1996, 61, 1994
and references therein.
(8) Yoshinari, K.; Yamazaki, K.; Komiyama, M. J . Am. Chem. Soc.
1991, 113, 5899.
S0022-3263(96)01178-4 CCC: $14.00 © 1997 American Chemical Society

Molecular Design for Pinpoint RNA Scission
J . Org. Chem., Vol. 62, No. 4, 1997 847
Sch em e 1
Sch em e 2
Resu lts
sizer. The coupling efficiencies for the steps involving 3
and 6 were 98 and 99%, respectively, as estimated from
the amount of released DMT. The benzyl and ethyl
esters remained intact during DNA synthesis.
Syn th esis of P h osp h or a m id ite Mon om er s 3 a n d
6. Phosphoramidite monomer 3 was synthesized accord-
ing to Scheme 1A. First, 2,2-bis(hydroxymethyl)propionic
acid was condensed with benzyl 4-aminobutylate using
1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydro-
chloride (EDCI). One of the two hydroxyl groups in 1
was protected by 4,4′-dimethoxytrityl chloride (DMT-Cl).
The final phosphitylation was made by 2-cyanoethyl
N,N,N′,N′-tetraisopropylphosphorodiamidite and 1H-tet-
razole. Another phosphoramidite monomer 6 was pre-
pared in a similar way, except for the use of glycine ethyl
ester and dicyclohexylcarbodiimide (DCC) in place of
benzyl 4-aminobutylate and EDCI, respectively (Scheme
1B). The resultant 3 and 6 were purified by column
chromatography and used for DNA synthesis.
Typical prepared oligonucleotides were 5′-TCA GCC
GGA TCA AXT CAT CGT-3′: 7-1 (by 3), 8-1 (by 6) and
5′-TCA GCC GGA TCX AGT CAT CGT-3′: 7-2 (by 3)
where X denotes the residue derived from 3 or 6.
Atta ch m en t of F u n ction a l Resid u es to Oligo-
n u cleotid es 7 a n d 8. Oligoamines were site-selectively
introduced to 7 and 8 by aminolysis of the benzyl or ethyl
ester (the first step in Scheme 2). The CPG columns used
for DNA synthesis were reacted directly with various
oligoamines in dry dioxane at 45 °C for 36 h. By this
simple method, the desired functionalized oligonucle-
otides, 11-Nm and 12-Nm , were obtained in high yields.
Here, m refers to the kind of oligoamine introduced. For
example, the product from 7-1 and diethylenetriamine
is designated as 11-1-N2, since an ethylenediamino
residue is bound to 7-1 via an amide linkage (note that
Oligon u cleotid e Syn th esis Usin g P h osp h or a m id -
ite Mon om er s 3 a n d 6. By using phosphoramidite
monomers 3 and 6, various oligonucleotides (7 and 8 in
Scheme 2) were prepared on an automated DNA synthe-

848 J . Org. Chem., Vol. 62, No. 4, 1997
Endo et al.
F igu r e 2. HPLC profiles for the oligonucleotide 11-1-N1 (A),
its enzymatically digested products (B), and authentic sample
of dAp(X-N1) (C); the enzymatic digestion was achieved by
alkaline phosphatase and snake-venom phosphodiesterase.
one of the terminal amino residues of the diethylenetri-
amine is used to form the amide linkage).
In a similar way, a cystamine residue was introduced
to 7 and 8. These cystamine-attached oligonucleotides
(11-Cys and 12-Cys) were further converted to the
corresponding thiol-attached ones (11-SH and 12-SH) by
reducing the disulfide linkage with dithiothreitol (DTT)
at pH 8.9
F igu r e 3. Autoradiographs of 20% denaturing polyacrylamide
gel electrophoresis for the hydrolysis of 28-mer RNA (32P-
labeled at the 5′-end) by 11-1-Nm (A) and by 11-2-Nm (B). A:
lane 1, ribonuclease T1 (G-specific); lane 2, no treatment; lane
3, alkaline hydrolysis; lane 4, 11-1-N1; lane 5, 11-1-N2; lane
6, 11-1-N3. B: lane 1, no treatment; lane 2, ribonuclease T1;
lane 3, alkaline hydrolysis; lane 4, 11-2-N1; lane 5, 11-2-N2;
lane 6, 11-2-N3. Reaction conditions: 0.2 mM RNA, 20 mM
artificial ribonuclease, 37 °C, 16 h, 50 mM Tris buffer (pH 7.5
for A and 8.5 for B), 10 mM EDTA. The cleavage efficiencies
increased with the reaction time and at 16 h were 3% for 11-
1-N3 (lane 6 in A) and 1% for 11-2-N3 (lane 6 in B).
When 11-1-N1 was digested by alkaline phosphatase
and snake-venom phosphodiesterase, dAp(X-N1), in
which an amino residue is bound to X in dApX via an
amide linkage, was formed together with the common
four deoxynucleosides (Figure 2b). The ratio (dA:dC:dG:
T:dAp(X-N1) ) 4:6:4:5:1) was exactly identical to the
theoretical one. The peak for dAp(X-N1) could be
concretely assigned by coinjection with its authentic
sample, which was prepared independently. The phos-
phodiester linkage in dAp(X-N1) was resistant to enzy-
matic digestion. Furthermore, a trimer dApXpT was
reacted with triethylenetetramine under the conditions
employed for the aminolysis of the oligonucleotides, and
the product was analyzed by mass spectroscopy. As
expected, the signals for dAp(X-N3)pT were observed (see
the Experimental Section).
Site-Selective RNA Scission by Oligon u cleotid es
11-1-N3 a n d 11-2-N3. Figure 3A shows the polyacryl-
amide gel electrophoresis pattern for RNA hydrolysis by
11-1-Nm . The two DNA portions in these conjugates are
complementary to the substrate 28-mer RNA at the A3-
A9 and the U11-A23 parts, respectively. When 11-1-N3
was used, the scission selectively occurred at the C10-
F igu r e 4. Scission profiles for the RNA hydrolysis by the
artificial ribonucleases 11-1-N3 and 11-2-N3; X denotes the
residue carrying an N3 residue. The scission sites are shown
by the arrows.
duplexes (see the result of the molecular-modeling study
presented in Figure 6A).
RNA hydrolysis by another artificial ribonuclease
11-2-N3 also took place selectively in front of the
diethylenetriamino residue (see lane 6 in Figure 3B
and Figure 4, as well as Figure 6B). In contrast, none of
U
11 and the U11-U12 linkages (lane 6: the scission profile
is presented in Figure 4). These scission sites were
located just in front of the diethylenetriamino residue,
when 11-1-N3 and the substrate RNA formed hetero-
(9) Fidanza, J . A.; McLaughlin, L. W. J . Org. Chem. 1992, 57, 2340.

Molecular Design for Pinpoint RNA Scission
J . Org. Chem., Vol. 62, No. 4, 1997 849
F igu r e 5. CD spectra of duplexes of the modified oligonucleotides 11-1-N3 (A) and 11-2-N3 (B) with the complementary RNA.
The concentration of the complementary RNA was varied as 0, 1, 2, and 3 mM (from the bottom), with the concentrations of the
modified oligonucleotide kept constant at 3 mM. The dotted lines are for the duplexes between the unmodified DNA (3 mM) and
the complementary RNA (3 mM). Measurement conditions are presented in the Experimental Section.
Ta ble 1. Meltin g Tem p er a tu r es (Tm , °C) of th e Du p lexes
the 11-1-N1, 11-1-N2, 11-2-N1, and 11-2-N2 was active
betw een th e Mod ified Oligon u cleotid es a n d Th eir
Com p lem en ta r y DNA a n d RNAa ,b
for RNA hydrolysis (lanes 4 and 5 for both A and B in
Figure 3).10 The N3 residues are essential for RNA
scission.
modified oligonucleotide
complementary
strand
Z
11-1-Z
11-2-Z
12-1-Z
Str u ctu r es of th e Du p lexes of 11-1-N3 a n d 11-2-
N3. The circular dichroism (CD) spectra of the duplexes
of modified oligonucleotides 11-1-N3 and 11-2-N3 with
the complementary RNA are presented in Figure 5.
Apparently, all the heteroduplexes take typical A-type
conformations in a wide range of molar ratio of the
modified DNA to the RNA.11 Furthermore, the spectra
for the heteroduplexes of the modified oligonucleotides
are nearly identical with the ones for the corresponding
duplexes between the unmodified DNA and the comple-
mentary RNA (the dotted lines in Figure 5). The chemi-
cal modification of the oligonucleotides does not cause any
drastic distortion of the duplexes with the complementary
RNA.
N1
N2
N3
Cys
SH
53.3
54.2
55.4
56.3
50.7c
55.8
57.1
57.3
53.1
54.2
54.4
54.5
52.9c
DNA
RNA
N3
52.1d
55.4d
a
The complementary strand is 5′-GCTAAACGATGACTTGATC-
CGGCTGATCGTC-3′ for DNA/DNA complexes and 5′-AAAAC-
GAUGACUUGAUCCGGCUGAAAAAAAAC-3′ for DNA/RNA com-
plexes. Tm’s were measured in pH 7 solutions containing 1 µM
DNA (or RNA), 10 mM sodium cacodylate, and 100 mM NaCl.
b
Measurement error is estimated to be <0.5 °C. Tm of the
unmodified DNA/DNA duplex is 65.6 °C, and the corresponding
value for the DNA/RNA complex is 63.3 °C. c 0.1 mM DTT was
d
added to the solutions. 1 mM EDTA was added to the solutions.
Th er m a l Sta bilities of th e Du p lexes of th e Mod i-
fied Oligon u cleotides. The melting temperatures (Tm’s)
of the duplexes between the modified oligonucleotides,
prepared in the present study, and either the comple-
mentary DNA or RNA are summarized in Table 1. In
all cases, the Tm’s are higher than 50 °C. Thus, the
duplexes can be satisfactorily formed under the physi-
ological conditions.
site (Figure 6B). Selective scission by the artificial
enzymes (Figures 3 and 4) occurs at the positions where
the N3 residue is accessible without causing much
instability in the duplexes.12
Discu ssion
Molecu la r Mod elin g on th e Heter od u p lexes of
Ar tificia l Ribon u clea ses w ith th e Su bstr a te RNA.
According to molecular modeling, the heteroduplex of
11-1-N3 with the RNA substrate is most stable when
the N3 residue (the catalytic site) is located near the
phosphodiester linkage between the C10 and the U11
residues (C10-p-U11) of the RNA; the corresponding
structure is depicted in Figure 6A, and the potential
energies are presented in Table 2. The placement of the
N3 residue at the U11-p-U12 linkage, another scission
site, also provides a stable structure. In the heteroduplex
of 11-2-N3 with the RNA, the U12-p-G13 linkage (the
selective scission site) is the most favorably accessible
In t er p osit ion of a n R NA Clea ver b et w een Tw o
DNA Oligom er s for a P in p oin t Scission . As clearly
evidenced above, an RNA cleaver, interposed between
two DNA oligomers, can be precisely placed at the
targeted phosphodiester linkage in substrate RNA. The
molecular movements of both the cleaver and the phos-
phodiester linkage are marginal in heteroduplexes. Un-
der these conditions, predicting the scission site is
(12) The calculation on the duplex of 11-2-N3 indicates that the
G13-p-A14 and A14-p-U15 linkages are more accessible by N3 than
the U11-p-U12 linkage (the latter linkage is another scission site).
Probably, structural changes of the strand around the X-residue cannot
be satisfactorily reproduced by the present calculation.
(13) Alternatively, the simultaneous protonation of both of the two
amino residues might be responsible for the lack of activities. Although
the pKa values of free ethylenediamine are 6.7 and 9.5, they can be
increased by negative charges of the DNA and RNA when the artificial
ribonucleases form heteroduplexes with the substrate RNA. Here, the
acid-base cooperation cannot operate.
(10) The selective hydrolysis was accompanied by some additional
scissions near the 3′-end. Assumedly, these linkages are activated for
alkaline hydrolysis on the hetero-duplex formation.
(11) Allen, F. S.; Gray, D. M.; Roberts, G. P.; Tinoco, I., J r.
Biopolymers 1972, 11, 853.

850 J . Org. Chem., Vol. 62, No. 4, 1997
Endo et al.
in high yields) incorporated into the desired sites of oligo-
nucleotides using aminolysis of the esters. According to
CD spectroscopy (Figure 5) and Tm measurements (Table
1), these modified oligonucleotides form stable duplexes
with RNA under the physiological conditions without any
significant structural distortion. Strong potentialities of
the present synthetic method have been indicated.
The incorporation of RNA cleaver between two DNA
oligomers can be also advantageous to achieve catalytic
turnover. When the substrate RNA is cleaved at the
inside of strand, the fragments are spontaneously re-
moved from the artificial ribonucleases since the melting
temperatures of their duplexes with the complementary
DNA portions are lower.
Mech a n ism of R NA H yd r olysis a n d Molecu la r
Mod elin g. The N3 residues in 11-1-N3 and 11-2-N3
are essential for RNA hydrolysis. The cleavage occurs
at the phosphodiester linkage, which the N3 can approach
favorably, as indicated by the molecular modeling. Prob-
ably, a neutral amino residue and an ammonium cation
show intramolecular acid-base cooperation.8 Under the
reaction conditions, two of the amino residues in N3 are
protonated and the other (the central one) is in a neutral
form.14 Consistently, 11-1-N2 and 11-2-N2 are inactive
(lanes 5 in Figure 3 A,B). It is sterically difficult to place
both of the two amino residues in the ethylenediamine
near the phosphodiester linkage.13 The argument is
further supported by the absence of the catalytic activi-
ties of 11-1-N1 and 11-2-N1 (lanes 4), since no intramo-
lecular acid-base cooperation takes place there.
Ethylenediamine tetraacetate (EDTA), a well-known
metal-ion chelator, showed no effect on selective scission.
The possibility that metal ions, if any, in the reaction
mixtures participate in the RNA hydrolysis is ruled out
(all the RNA hydrolyses in Figure 3 were achieved in the
presence of 10 mM EDTA). Note that contamination by
metal ions was carefully avoided by using highly purified
water (the specific resistance >18.3 MΩ) as the solvent.
F igu r e 6. Stereoviews of the most stable structures of the
duplexes of the substrate RNA with 11-1-N3 (A) and 11-2-
N3 (B): the bold line shows the side chain containing the
diethylenetriamine N3, whereas the hydrolyzed phosphodiester
linkage is indicated by the circle with an arrow. The calculation
was made using the Biosym Insight II/Discover program on
nine residues for the artificial ribonuclease and the RNA. The
assumptions employed for the calculation are presented in the
Experimental Section.
Ta ble 2. Tota l P oten tia l En er gy (k ca l/m ol) for th e
Heter od u p lexes betw een th e Mod ified Oligon u cleotid es
a n d th e Com p lem en ta r y RNA, in w h ich th e Ter m in a l
Am in o Resid u es in N3 a r e P la ced a t th e Cor r esp on d in g
P h osp h od iester Lin k a gea
Con clu sion
Phosphoramidite monomers 3 and 6 are very useful
for attaching various functional residues inside of DNA
strands. When these modified oligonucleotides form
duplexes with the complementary nucleic acids, the
positions of the functional residues are precisely con-
trolled, and thus, the reactions take place selectively at
the target sites. The site-selective RNA hydrolyses
exhibited in Figures 3 and 4 are typical examples of the
application. It is noteworthy that the scission sites are
identical with the sites predicted by a molecular-modeling
study. The scission efficiencies should be improved by a
more precise molecular design. These artificial enzymes
are valuable for molecular biology, biotechnology, therapy,
and others, since the scission sites are highly restricted
and predicted. These attempts are currently under way
in our laboratory.
11-1-N3
phosphodiester
11-2-N3
phosphodiester
potential
energy
potential
energy
linkage
linkage
A9-C10
-5597
-5797
-5677
-5642
C10-U11
U11-U12b
U12-G13b
G13-A14
-5674
-5724
-5921
-5859
C10-U11b
U11-U12b
U12-G13
a
Calculation was made for 9-mer heteroduplexes (d(ATCAAX-
TCA)/r(UGACUUGAU) for 11-1-N3 and d(GGATCXAGT)/r(ACU-
UGAUCC) for 11-2-N3) (see Experimental Section for details).
b
The hydrolyzed phosphodiester linkages.
straightforward. Undoubtedly, this is a great advantage
of the present artificial ribonucleases over the previous
ones (see Introduction) from the viewpoint of practical
applications.
Exp er im en ta l Section
Phosphoramidite monomers 3 and 6 are eminent for
preparing these artificial ribonucleases. Both of them
provide three carbon atoms to the backbone of the oligo-
nucleotides, as do the common phosphoramidite mono-
mers. A less bulky and flexible trimethylene residue is
used to minimize the steric hindrance. In addition to
oligoamines, various functional residues are easily (and
Ma ter ia ls. 2-Cyanoethyl N,N,N′,N′-tetraisopropylphospho-
rodiamidite and triethylenetetramine were obtained from
Aldrich. 2,2-Bis(hydroxymethyl)propionic acid, cystamine hy-
(14) The pKa values of diethylenetriamine are 4.6, 8.6, and 9.7, as
determined by potentiometric titration (protonation of the central
amine is greatly suppressed by the electrostatic repulsion by the
neighboring two positive charges).

Molecular Design for Pinpoint RNA Scission
J . Org. Chem., Vol. 62, No. 4, 1997 851
drochloride, DCC, 4-(dimethylamino)pyridine (DMAP), EDCI,
glycine ethyl ester hydrochloride, and 1H-tetrazole, as well as
ethylenediamine and diethylenetriamine, were purchased from
Tokyo Kasei. 3-Hydroxybenzotriazole (HOBt) was obtained
from Wako, and DMT-Cl was from Dojin. 1,4-Dioxane and
pyridine were distilled from metallic sodium and potassium
hydroxide, respectively. Acetonitrile, CH2Cl2, DMF, and the
oligoamines were dried over molecular sieves (4 Å). The
conventional phosphoramidite agents (Expedite) were pur-
chased from PerSeptive Biosystems. Silica gel column chro-
matography was achieved using Kiselgel 60 from Merck, and
TLC was performed on precoated silica gel plates (Merck
Kiselgel 60 F254). The enzymes used were alkaline phos-
phatase and snake-venom phosphodiesterase from Boehringer
Mannheim, and T4 polynucleotide kinase was from Nippon
Gene.
Eth yl 2-[N-[2,2-Bis(h yd r oxym eth yl)p r op ion yl]a m in o]-
eth yla te (4). Triethylamine (6.95 mL, 50 mmol) was added
to glycine ethyl ester hydrochloride (7.00 g, 50 mmol) in 25
mL of DMF, and the formed triethylammonium hydrochloride
was filtered out. To the filtrate was added 2,2-bis(hydroxy-
methyl)propionic acid (7.40 g, 55 mmol); then, DCC (11.4 g,
55 mmol) in 20 mL of dioxane was slowly dropped into this
mixture on ice. The mixture was stirred for 12 h at room
temperature and filtered again. The filtrate was concentrated
under reduced pressure and purified by silica gel column
chromatography (CH2Cl2:i-PrOH ) 100:5), giving 8.19 g of 4
(75% yield): 1H-NMR [CDCl3(TMS)] δ 7.68 (brt, 1H), 4.12 (q,
J ) 7.2 Hz, 2H), 4.05 (d, J ) 5.9 Hz, 2H), 3.79 (brs, 6H), 1.29
(t, J ) 7.2 Hz, 3H), 1.09 (s, 3H); Rf ) 0.11 (CH2Cl2:i-PrOH )
100:5).
E t h yl 2-[N-[2-[[(4,4′-Dim et h oxyt r it yl)oxy]m et h yl]-2-
(h yd r oxym eth yl)p r op ion yl]a m in o]eth yla te (5). A CH2-
Cl2 solution (10 mL) of 4 (5.58 g, 21.8 mmol), DMAP (0.12 g,
1.0 mmol), and N,N-diisopropylethylamine (5.2 mL, 30 mmol)
was cooled on ice under nitrogen, and DMT-Cl (6.78 g, 20
mmol) in 10 mL of CH2Cl2 was dropwise added. After 12 h,
the product was treated as for 2, and purified by silica gel
column chromatography (CH2Cl2:i-PrOH ) 100:2). The yield
was 4.80 g (43%): 1H-NMR [CDCl3(TMS)] δ 7.18-7.43 (m, 9H),
6.84 (d, J ) 8.9 Hz, 4H), 4.19 (q, J ) 7.3 Hz, 2H), 4.00 (dd, J
) 4.3, 5.6 Hz, 2H), 3.80 (s, 6H), 3.64 (d, J ) 6.6 Hz, 2H), 3.30
(dd, J ) 9.2, 23.4 Hz, 2H), 2.35 (t, J ) 7.3 Hz, 2H), 1.18 (s,
3H); Rf ) 0.39 (CH2Cl2:MeOH ) 95:5).
P h osp h or a m id ite Mon om er 6. In 10 mL of dry acetoni-
trile, 5 (0.52 g, 1.0 mmol) and 2-cyanoethyl N,N,N′,N′-
tetraisopropylphosphorodiamidite (0.35 mL, 1.1 mmol) were
treated with 1H-tetrazole (77 mg, 1.1 mmol) under nitrogen
for 1 h. After the usual workup, the mixture was dried over
Na2SO4 and purified by silica gel column chromatography
(hexane:AcOEt ) 3:1) to give 0.33 g of the phosphoramidite 6
(52% yield): 1H-NMR [CDCl3(TMS)] δ 7.18-7.43 (m, 9H), 6.82
(d, J ) 8.9 Hz, 4H), 4.17 (q, J ) 6.9 Hz, 2H), 3.96 (q, J ) 5.3
Hz, 2H), 3.79 (s, 6H), 3.66-3.75 (m, 4H), 3.55 (m, 2H), 3.29
(m, 2H), 2.54 (m, 2H), 1.08-1.28 (m, 18H); 31P-NMR (CDCl3)
δ 154.03, 154.07; Rf ) 0.54 (hexane:AcOEt ) 1:1).
The substrate 28-mer RNA was prepared on a DNA syn-
thesizer and purified in the recommended fashion.15 The
choice of sequence was arbitrary. The RNA was labeled at
the 5′-end by adenosine [γ-32P]triphosphate (from Amersham)
with T4 polynucleotide kinase and purified by polyacrylamide
gel electrophoresis.
Ben zyl4-[N-[2,2-Bis(h yd r oxym eth yl)p r op ion yl]a m in o]-
bu tyla te (1). EDCI (3.00 g, 15.6 mmol) was added to a
mixture of benzyl 4-aminobutylate (2.99 g, 13.0 mmol), 2,2-
bis(hydroxymethyl)propionic acid (2.09 g, 15.6 mmol), HOBt
(2.12 g, 15.6 mmol), triethylamine (2.72 mL, 19.5 mmol), and
DMF (10 mL). After the mixture was stirred for 12 h at room
temperature, the solvent was removed, and 100 mL of ethyl
acetate was poured. The organic layer was washed with 100
mL of saturated aqueous solutions of NaHCO3 (once) and of
NaCl (twice), dried over anhydrous Na2SO4, and concentrated
under reduced pressure. The product was purified by silica
gel column chromatography (CH2Cl2:MeOH ) 95:5) to afford
2.84 g of 1 (63% yield): 1H-NMR [CDCl3(TMS)] δ 7.33 (brs,
5H), 7.02 (brs, 1H), 5.10 (s, 2H), 3.67 (d, J ) 4.0 Hz, 4H), 3.26
(q, J ) 5.9 Hz, 2H), 3.07 (t, J ) 5.0 Hz, 2H), 2.39 (t, J ) 7.6
Hz, 2H), 1.84 (quin, J ) 6.9 Hz, 2H), 1.05 (s, 3H); Rf ) 0.28
(CH2Cl2:MeOH ) 10:1).
Ben zyl 4-[N-[2-[[(4,4′-Dim eth oxytr ityl)oxy]m eth yl]-2-
(h yd r oxym eth yl)p r op ion yl]a m in o]bu tyla te (2). A dry
pyridine solution (10 mL) of 1 (2.84 g, 8.22 mmol) and DMAP
(53 mg, 0.41 mmol) was cooled on ice under nitrogen, and
DMT-Cl (3.34 g, 9.86 mmol) in 5 mL of CH2Cl2 was added
dropwise. After 12 h, 100 mL of CH2Cl2 was poured into the
mixture. The CH2Cl2 layer was dried with anhydrous Na2-
SO4 after being washed with saturated aqueous solutions of
NaHCO3 (twice) and of NaCl (once). Silica gel column chro-
matography (CH2Cl2:i-PrOH:Et3N ) 97.5:2.5:1) gave 4.87 g of
2 (59% yield): 1H-NMR [CDCl3(TMS)] δ 7.20-7.41 (m, 14H),
7.04 (brt, 1H), 6.83 (d, J ) 8.6 Hz, 4H), 5.09 (s, 2H), 3.77 (s,
6H), 3.59 (d, J ) 6.6 Hz, 2H), 3.20-3.32 (m, 4H), 2.35 (t, J )
7.6 Hz, 2H), 1.81 (quin, J ) 7.2 Hz, 2H), 1.19 (s, 3H); Rf )
0.19 (CH2Cl2:i-PrOH:Et3N ) 97.5:2.5:1).
P h osp h or a m id ite Mon om er 3. In a dry acetonitrile
solution (10 mL), 2 (0.28 g, 0.45 mmol) and 2-cyanoethyl
N,N,N′,N′-tetraisopropylphosphorodiamidite (0.16 mL, 0.50
mmol) were reacted with 1H-tetrazole (35 mg, 0.50 mmol)
under nitrogen. Prior to the reaction, 2 and 1H-tetrazole were
dried by coevaporation with dry acetonitrile (twice). After 1
h, the product was taken into 100 mL of ethyl acetate. The
organic solution was washed with 100 mL of saturated aqueous
solutions of NaHCO3 and of NaCl, dried over Na2SO4, and
desalted; finally, the solvent was removed in vacuo. The
resultant product was purified by silica gel column chroma-
tography (hexane:AcOEt ) 3:1) to afford 0.22 g of phosphora-
midite 3 (67% yield): 1H NMR [CDCl3(TMS)] δ 7.19-7.44 (m,
14H), 6.82 (d, J ) 8.6 Hz, 4H), 5.08 (s, 2H), 4.11 (q, J ) 6.9
Hz, 2H), 3.76 (s, 6H), 3.68 (m, 2H), 3.51 (m, 2H), 3.24 (m, 4H),
2.73 (m, 1H), 2.48 (q, J ) 5.9 Hz, 2H), 2.33 (t, J ) 7.6 Hz,
2H), 1.76 (quin, J ) 7.6 Hz, 2H), 1.07-1.29 (m, 15H); 31P-NMR
(CDCl3) δ 152.37, 152.41; Rf ) 0.59 (hexane:AcOEt ) 1:1).
Syn th esis of Mod ified Oligon u cleotid es 7 a n d 8. By
using phosphoramidite monomers 3 and 6, together with the
conventional Expedite agents, various oligodeoxyribonucle-
otides were automatically synthesized on a MilliGen/Biosearch
Cyclone Plus DNA synthesizer. The concentrations (52 mM)
of 3 and 6 in dry acetonitrile were 1.5 times as large as those
used for the Expedite agents.
Atta ch m en t of Am in o Com p ou n d s to 7 a n d 8. After
DNA synthesis, the CPG column was kept in 1.5 mL of a 1:2
mixture of the amine and dioxane at 45 °C for 36 h and then
treated with concentrated NH4OH for 1 h at room temperature.
The oligonucleotide (with the DMT on) was purified by
reversed-phase HPLC and further treated with a 4:1 mixture
of acetic acid and acetonitrile for 1 h. The final purification
was also carried out by HPLC (conditions: 5-20% acetonitrile/
water (20 min), 50 mM ammonium formate, 260 nm, 1.0 mL/
min).
In order to both measure the concentrations of these
oligonucleotides and to confirm their structures, they were
hydrolyzed by alkaline phosphatase and snake-venom phos-
phodiesterase at pH 7.0 and 37 °C for 2 h. The digests were
analyzed by reversed-phase HPLC: 5-10% acetonitrile/water
(20 min) for 11-1-Nm , 11-2-Nm and 12-1-Nm , and 5-20%
(20 min) for the others.
P r ep a r a t ion of a n Au t h en t ic Sa m p le of d Ap (X-N1).
The Expedite agent for dA (N6-(tert-butylphenoxyacetyl)-5′-O-
(4,4′-dimethoxytrityl)-2′-deoxyadenosine 3′-O-(2-cyanoethyl N,N-
diisopropylphosphoramidite)) (100 mg, 0.106 mmol) and 2 (65
mg, 0.106 mmol) were reacted with 1H-tetrazole (24 mg, 0.32
mmol) in 10 mL of acetonitrile under nitrogen for 1 h at room
temperature. Then, iodine (40 mg, 0.16 mmol) in 5 mL of THF
and 1 mL of 2,6-lutidine were added to the mixture. After 30
min, the mixture was concentrated and dissolved in 100 mL
(15) Gait, M. J .; Pritchard, C.; Slim, G. Oligonucleotides and
Analogues; Eckstein, F., Ed.; IRL Press: Oxford, 1991; p 25.

852 J . Org. Chem., Vol. 62, No. 4, 1997
Endo et al.
of CH2Cl2. The organic layer was washed with 100 mL of
saturated aqueous solutions of Na2S2O3 and of NaHCO3 and
dried over Na2SO4. Silica gel column chromatography (CH2-
Cl2:MeOH:Et3N ) 100:10:1) afforded the fully protected dApX
in 25% yield (38 mg).
Sequ en ce-Selective Hyd r olysis of RNA by th e Oli-
goa m in e-Atta ch ed Oligon u cleotid es. 5′-32P-end-labeled
28-mer RNA (0.2 mM) was reacted in Tris buffers (50 mM)
containing 20 mM artificial ribonuclease and 10 mM EDTA.
After 16 h at 37 °C, the reaction mixture was analyzed on 20%
denatured polyacrylamide gel. Quantification of RNA cleavage
was carried out on a Fujix BAS 1000 bioimaging analyzer.
Molecu la r Mod elin g Stu d y. Optimized modeling on
duplexes of the artificial enzymes and the substrate RNA was
performed using Biosym Insight II/Discover software. The
following assumptions were made: (1) The heteroduplexes
have A-type conformations (see the CD spectra in Figure 5).
(2) The central carbon atoms of the trimethylene residues,
derived from 3, take S-type configurations. (3) The tethers
are accommodated in the minor groove. (4) The distance
between the nonbonding oxygen of the phosphodiester linkage
and the primary nitrogen of N3 is restrained at 3.0 Å, whereas
the distance between the hydrogen in the 2′-OH of the ribose
(in the 5′-side of the scission site) and the secondary nitrogen
of N3 is kept constant at 1.8 Å. The fourth assumption was
made in terms of the catalytic mechanism in which one of the
amino residues abstracts a proton from the 2′-OH and the
adjacent ammonium ion acts as a general acid catalyst.8
Under these assumptions, the minimum energies (1000 steps)
were calculated for each of the phosphodiester linkages of the
substrate RNA, where a 5 Å layer of water molecules on the
surface of the duplex was taken into consideration.
The protected dApX was incubated in 1.5 mL of a 1:2
mixture of ethylenediamine and dioxane at 40 °C for 48 h. The
product was successively treated with 29% NH4OH for 1 h at
room temperature and then with 4:1 acetic acid-acetonitrile
solution for 1 h. The product was dissolved in water, filtered,
and purified by HPLC [5-20% acetonitrile/water (20 min)
containing 50 mM ammonium formate]: 1H-NMR [D2O (TSP)]
δ 8.34 (s, 1H), 8.25 (s, 1H), 6.51 (t, J ) 6.3 Hz, 1H), 4.37 (q, J
) 2.3 Hz, 1H), 4.01 (m, 1H), 3.84-3.90 (m, 4H), 3.77 (d, J )
11.0 Hz, 1H), 3.62 (d, J ) 11.0 Hz, 1H), 3.48 (t, J ) 5.6 Hz,
2H), 3.24 (q, J ) 6.3 Hz, 2H), 3.15 (q, J ) 5.9 Hz, 2H), 2.84-
2.90 (m, 1H), 2.73-2.76 (m, 1H), 2.28 (t, J ) 7.6 Hz, 2H), 1.77
(t, J ) 7.6 Hz, 2H), 1.24 (s, 3H); electrospray ionization mass
spectroscopy (ESI-MS) m/z 573.9 [M - H]-.
P r ep a r a tion of th e P r od u ct of d Ap Xp T w ith Tr ieth -
ylen etetr a m in e by Solid -P h a se Syn th esis. An authentic
sample of dApXpT, prepared on a DNA synthesizer, was kept
in 1.5 mL of a 1:2 mixture of triethylenetetramine, and dioxane
at 45 °C for 36 h. After being treated with 29% NH4OH for 1
h and concentrated, the product was purified by reversed-
phase HPLC: ESI-MS m/z 963.3 [M - H]-, 965.3 [M + H]+,
987.3 [M + Na]+.
Melt in g Tem p er a t u r e Mea su r em en t s. The melting
profiles of the duplexes were obtained at 260 nm in a quartz
cell of 1 cm path length fitted with a Teflon stopper. The
heating rate was 0.5 °C/min. The specimens contained modi-
fied oligonucleotide (1 µM), complementary DNA (1 µM), and
NaCl (100 µM) in pH 7.0 cacodylate buffers (10 mM). In the
case of 11-1-SH and 12-1-SH, 0.1 mM DTT was added to
the solutions in order to avoid oxidation of the thiol residues.
For the DNA/RNA heteroduplexes, 1 mM EDTA was added.
The experimental error in Tm was estimated to be within 0.5
°C.
Sp ectr oscop y. CD spectra were obtained on a J ASCO
J -500A spectropolarimeter at 15 °C in pH 8 solutions contain-
ing modified DNA (3 µM), complementary 32 mer RNA (0, 1,
2, and 3 µM), Tris-HCl (10 mM), NaCl (100 mM), and EDTA
(1 mM). 1H-NMR spectra (with TMS or TSP as internal
standard) and 31P-NMR spectra (85% H3PO4 in D2O as external
standard) were recorded on a J EOL EX-270 NMR spectrometer
at 270 and 109.4 MHz, respectively.
Ack n ow led gm en t. The authors are grateful to
J apan Spectroscopic Co., Ltd. (J ASCO, Tokyo) for
measuring the mass spectra. This work was partially
supported by the Ministry of Education, Science and
Culture: Large-scale Research Project under the New
Program System in Grants-in-Aid for Scientific Re-
search. The award of a J SPS Research Fellowship for
Young Scientists to M.E. is also acknowledged.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
1-6 and dAp(X-N1), 31P-NMR spectra for 3 and 6, and mass
spectra for dAp(X-N1) and dAp(X-N3)pT (6 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O9611780