3-Deazaadenosine analogues of p5′A2′p5′A2′p5′ A: Synthesis, stereochemistry, and the roles of adenine ring nitrogen-3 in the interaction with RNase L

Article abstract of DOI:10.1016/j.bmc.2004.04.021

Sequence-specific 3-deazaadenosine (c³A)-substituted analogues of trimeric 2′,5′-oligoadenylate, p5′A2′p5′ A2′p5′A, were synthesized and evaluated for their ability to activate human RNase L (EC 3.1.2.6) aiming at the elucidation of the nitrogen-3 role in this biochemical process. Substitution of either 5′-terminal or 2′-terminal adenosine with c³A afforded the respective analogues p5′(c³A)2′p5′A2′p5′A and p5′A2′p5′A2′p5′(c³A) that were as effective as the natural tetramer itself as activators of RNase L (EC ₅₀=1nM). In contrast, p5′A2′p5′(c ³A)2′p5′A showed diminished RNase L activation ability (EC₅₀=10nM). The extensive conformational analysis of the c ³A-substituted core trimers versus the parent natural core trimer by the ¹H and ¹³C NMR, and CD spectroscopy displayed close stereochemical similarity between the natural core trimer and (c ³A)2′p5′A2′p5′A and A2′p5′ A2′p5′(c³A) analogues, thereby strong evidences for the syn base orientation about the glycosyl bond of the c³A residue of the latter were found. On the contrary, an analogue A2′p5′(c ³A)2′p5′A displayed rather essential deviations from the spatial arrangement of the parent natural core trimer.

Full text of DOI:10.1016/j.bmc.2004.04.021

Bioorganic & Medicinal Chemistry 12 (2004) 3637–3647
3-Deazaadenosine analogues of p50A20p50A20p50A: synthesis,
stereochemistry, and the roles of adenine ring nitrogen-3 in the
interaction with RNase L
Elena N. Kalinichenko,a Tatiana L. Podkopaeva,a Elena V. Budko,a Frank Seela,b
c
Beihua Dong, Robert Silverman, Jouko Vepsalainen, Paul F. Torrence and
c
d
e
€ €
Igor A. Mikhailopuloa,f,
*
aInstitute of Bioorganic Chemistry, National Academy of Sciences, 220141 Minsk, Belarus
bInstitut f€ur Chemie, Universit€at Osnabr€uck, D-49069 Osnabr€uck, Germany
cCancer Biology, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH 44195, USA
dDepartment of Chemistry, University of Kuopio, FIN-70211 Kuopio, Finland
eDepartment of Chemistry, Northern Arizona University, Flagstaff, AZ 86011-5698, USA
fDepartment of Pharmaceutical Chemistry, University of Kuopio, PO Box 1627, FIN-70211 Kuopio, Finland
Received 11 February 2004; accepted 15 April 2004
Available online 18 May 2004
Abstract—Sequence-specific 3-deazaadenosine (c3A)-substituted analogues of trimeric 20,50-oligoadenylate, p50A20p50A20p50A, were
synthesized and evaluated for their ability to activate human RNase L (EC 3.1.2.6) aiming at the elucidation of the nitrogen-3 role in
this biochemical process. Substitution of either 50-terminal or 20-terminal adenosine with c3A afforded the respective analogues
p50(c3A)20p50A20p50A and p50A20p50A20p50(c3A) that were as effective as the natural tetramer itself as activators of RNase L
(EC50 ¼ 1 nM). In contrast, p50A20p50(c3A)20p50A showed diminished RNase L activation ability (EC50 ¼ 10 nM). The extensive
conformational analysis of the c3A-substituted core trimers versus the parent natural core trimer by the 1H and 13C NMR, and CD
spectroscopy displayed close stereochemical similarity between the natural core trimer and (c3A)20p50A20p50A and
A20p50A20p50(c3A) analogues, thereby strong evidences for the syn base orientation about the glycosyl bond of the c3A residue of the
latter were found. On the contrary, an analogue A20p50(c3A)20p50A displayed rather essential deviations from the spatial
arrangement of the parent natural core trimer.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
thus inhibits protein synthesis. Moreover, numerous
biochemical studies of 2-5A manifested a broad palette
of processes that are influenced and/or regulated by
these oligomers.2
The 20,50-linked 50-O-phosphorylated oligoadenylates (2-
5A) play a key role in the antiviral action of inter-
feron.1;2 In the presence of double-stranded RNA,
interferon induces in vertebrate cells the production of
the enzyme (20,50)oligoadenylate synthetase, which uti-
lize ATP to generate a unique group of 20,50-phospho-
diester-linked oligomers referred to as 2-5A,
(pp)p50A20(p50A20)np50A (n ¼ 1–3; mainly trimer, n ¼ 1)
(Scheme 1). These oligomers bind to and subsequently
activate RNase L (EC 3.1.2.6), a constitutive, but latent
endonuclease that degrades the mRNA of the virus and
Since the exact chemical structure of 2-5A was estab-
lished,1 a number of analogues of 2-5A have been syn-
thesized in order to examine the crucial structural and
stereochemical parameters of 2-5A for binding to and
activation of RNase L. It was established that diverse
functionalities of each individual nucleotide fragment of
2-5A contribute highly specifically to binding to and
activation of RNase L. Thus, the role of each purine N7-
atom in both processes was studied with a set of trimeric
2-5A analogues, in which one or more adenosine resi-
dues was replaced by 7-deazaadenosine (c7A; tuberci-
din).3;4 Evaluation of c7A analogues of 2-5A for their
ability to bind to and activate RNase L of mouse L cells
Keywords: 2-5A; Analogues; 3-Deazaadenosine; RNase L.
* Corresponding author. Tel.: +358-17-163-128/+358-40-085-1055;
fax: +358-17-162-456; e-mail: igor.mikhailopulo@uku.fi
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.04.021

3638
E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
NH2
N
O
O
N
P
A1
O
NH2
RO
O
N
X
N
N
N
O
A2
O
NH2
HO
O
P
O
Y
O
N
N
N
O
A3
HO
O
P
O
Z
O
O
HO
OH
2-5A, X = Y = Z = N; R = phosphate or pyrophosphate
1, X = CH; Y = Z = N; R = H
2, Y = CH; X = Z = N; R = H
3, Z = CH; X = Y = N; R = H
NH2
4(6)
3a(5)
7a(4)
3(7)
1(9)
N
N
5(1)
N
6(2)
HO
C
H
O
7(3)
HO
4; c3A
Systematic numbering; purine numbering in parenthesis
OH
Scheme 1.
showed that there were small changes (6 10-fold) in
their ability to bind to RNase L. On the contrary, an
essential decrease in the ability to activate RNase L was
found upon replacement of either the 50-terminal (A1,
33-fold) or 20-terminal (A3, 7-fold) adenosine of 2-5A by
c7A. Only one analogue, (pp)p50A20p50(c7A)20p50A, in
which the central adenosine residue A2 was replaced by
c7A, retained activity equivalent to 2-5A itself.3;4
diminution in affinity to RNase L, respectively. The 50-
terminus (A1) c1A-substituted analogue, p50(c1A)20p50
A20p50A, showed a 35-fold reduction in RNase L acti-
vation ability. The congeners with the A2 or A3 aden-
osine substitution displayed a slight drop in RNase
L activation ability, viz., the 2-fold and 4-fold diminu-
tion, respectively. Under similar conditions, the inosine-
substituted analogues p50I20p50A20p50A and p50A20p50-
A20p50I were found to be a 31-fold and 142-fold less
efficient in RNase L activation ability. Again, an
analogue p50A20p50(c1A)20p50A, in which the central
adenosine residue was replaced by c1A, retained activity
equivalent to 2-5A itself. Thus, the exocyclic amino
group of the 20-terminal adenosine A3 is critical for
RNase L activation.7
Replacement of the adenine base of 2-5A by hypoxan-
thine allowed to appreciate the role of N6-amino/N1
functionality in the interaction with RNase L of mouse
cells.5;6 The exchange at the 50-terminus of 2-5A led to a
large decrease both in RNase L binding affinity and in
activation ability (200-fold). Similarly, adenine fi
hypoxanthine exchange at the 20-terminus resulted in
dramatic 1000-fold drop in activation ability with little
change in RNase L binding ability. On the contrary, the
adenine fi hypoxanthine replacement in the middle led
to decrease in RNase L activation (20-fold loss) without
essential deterioration of binding ability (2–3-fold de-
crease). A further refinement of the role of N6-amino/N1
functionality in the interaction with the recombinant
human RNase L was made recently by us employing a
set of 2-5A analogues, in which adenosine residues have
been successively replaced by 1-deazaadenosine (c1A).7
It was established that the nitrogen-1 of the 50-terminal
adenosine moiety A1 is key for binding to RNase L and
its replacement by c1A gives rise to a 81-fold decrease
in binding affinity. Substitution of A2 or A3 adenine
residues with c1A resulted in a 9-fold and 15-fold
In order to gain further insight into the role of the adenine
nitrogen atoms in binding to and activation of RNase L,
we have synthesized 2-5A analogues 1–3 with consecutive
replacement of adenosine residues with 3-deazaadenosine
(c3A; 4) and evaluated their ability to bind to and activate
human recombinant RNase L (EC 3.1.2.6).
2. Results and discussion
2.1. Synthesis of 3-deazaadenosine oligonucleotides 20–22
and their 50-monophosphates 1–3
They were synthesized essentially as described previ-
ously7;8 by the use a phosphotriester methodology

E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
3639
applying the adenosine and 3-deazaadenosine 20-termi-
nal and phosphodiester building blocks 12/13 and 9/11,
respectively. It was previously shown that the 30-O-
benzoyl protection in combination with 50-O-mono-
methoxytritylation and the 2-(4-nitrophenyl)ethyl
(NPE) group for phosphate protection is effective for
(20,50)oligonucleotide synthesis.9–11 By the analogy with
the synthesis of 1-deazaadenosine analogues of 2-5A,7;8
2-(4-nitrophenyl)ethoxycarbonyl (NPEOC) group12 was
successfully employed for the protection of amino
function of c3A. Transient protection protocol13 of hy-
droxyl groups of c3A with TMS followed by the reaction
with NPEOC chloride gave, after work-up and silica gel
column chromatography, compound 5 in 68% yield.
Treatment of the NPEOC derivative 5 with DBU in
pyridine at room temperature for 2 h resulted in com-
plete deprotection affording c3A as the only reaction
product (Scheme 2).
The assembly of the trimers was performed by con-
densing [TPS-Cl/N-methylimidazole, molar ratio 1:3, as
an activating agent; CHCl3 as a solvent (cf., e.g., Ref. 8)]
the monomeric building blocks in different successions
and combinations in order to synthesize the 50-detrityl-
ated dimers 14–16. These latter dimers were condensed
with the phosphodiesters 11 or 13 followed by detrityl-
ation to afford the partially blocked trimers 17–19,
which were used for the synthesis of the core c3A
oligonucleotides 20–22 as well as their 50-monophos-
phates 1–3. Deprotection of 17–19 was performed by the
treatment at room temperature with DBU in pyridine
for 2 h followed by saturated methanolic ammonia for
20 h to give, after DEAE-Sephadex A-25 (HCO3ꢀ-form)
chromatography, the corresponding trimers 20–22 in
44–52% overall yield (see Section 4 and Table 1). On the
other hand, the treatment of the 50-detritylated trimers
17–19 with pyrophosphoryl chloride in ethyl acetate9;14
followed by deprotection and purification as just de-
scribed for the core trimers afforded the desired 50-
monophosphates 1–3 as amorphous Naþ salts in 35–
49% overall yield (see Section 4 and Table 1) (Scheme 3).
The method of selective 30-O-benzoylation9 using freshly
distilled benzoyl chloride (1.065 equiv) in acetonitrile in
the presence of Et3N and DMAP was now carried out
on compound 6 to give the respective 30-O-benzoylated
derivative 8 in a yield of 60%, along with 20,30-di-O-
benzoylated by-product 7 (13%). The former was
transformed to the corresponding phosphotriester 10,
and then to the phosphodiester 11 under conventional
conditions.10 Benzoylation of 6 and subsequent detrityl-
ation gave the 20-terminal c3A building block 9. The
synthesis of analogous adenosine building blocks 12 and
13 was previously described.10
2.2. Conformational studies
The main goal of these studies was to give further
insight into the role of structural and/or stereochemical
factors in defining the biochemical properties of (20-
50)oligoadenylates. The conformation of the core trimers
1
20–22 in aqueous solution was studied by H and 13C
NHX
NYZ
N
N
N
N
N
N
R1O
R1O
C
N
O
O
H
R2O
OR3
R2O
OR3
5, R1 = R2 = R3 = H
12, R1 = H; Y = Z = R2 = R3 = Bz
6, R1 = MTr; R2 = R3 = H
13, R1 = MTr; Y = R2 = Bz; Z = H
7, R1 = MTr; R2 = R3 = Bz
8, R1 = MTr; R2 = Bz; R3 = H
9, R1 = H; R2 = R3 = Bz
O
R3 =
P
ONPE
O NEt3H
10, R1 = MTr; R2 = Bz;
O
R3 =
P
ONPE
OoClPh
11, R1 = MTr; R2 = Bz;
O
R3 =
P
ONPE
O NEt3H
NO2
NO2
O
O
X =
NPE =
(NPEOC)
Scheme 2.

3640
E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
Table 1. Isolated yields, hypochromicity, and HPLC data of the 3-deazaadenosine core analogues of (20,50)A3 20–22 and their 50-monophosphates
1–3
Compd
Isolated yield [puritya] (%)
Hypochromicityb (%)
Retention time (min)
ApAp(c3A) (20)
Ap(c3A)pA (21)
(c3A)pApA (22)
p(c3A)pApA (1)
pAp(c3A)pA (2)
pApAp(c3A) (3)
48 [86.1]
52 [88.4]
44 [86.6]
40 [95.7]
35 [93.5]
49 [69.4]
25.0
21.0
14.5
––
5.72
5.35
5.38
8.97
––
––
11.28
12.08
a According to HPLC data for compounds after chromatography (see Section 4); all biochemical and spectral experiments have been performed with
compounds that were purified to homogeneity by reverse phase HPLC (purity P97%).
b Hypochromicity was determined as described previously.8
NHR2
N
N
NHR3
HO
N
X
O
N
N
N
O
NHR4
R1O
O
P
O
Y
O
O
OR
N
N
N
O
R1O
O
P
Z
O
OR
n
R1O
OR1
R4
n = 0
No.
X
Y
Z
R
R1
R2
R3
14
15
16
N
N
CH
-
-
-
N
NPE Bz
CH NPE Bz
NPE Bz
Bz
Bz
NPEOC -
-
-
Bz
NPEOC
Bz
N
n = 1
17
18
19
20
21
22
N
N
CH
N
N
N
CH
N
N
CH
N
CH NPE Bz
Bz
Bz
Bz
NPEOC
N
N
NPE Bz
NPE Bz
NPEOC Bz
NPEOC Bz
Bz
H
H
CH H
H
H
H
H
H
H
H
H
H
N
N
H
H
CH
H
Scheme 3.
3
0
NMR spectroscopy (Tables 2–5) and CD spectroscopy
(Fig. 1). The assignment of all NMR resonances was
made by 2D [1H,1H] and [1H,13C] correlation spectra.
2.5–3.7 Hz along with the JC8;H1 ¼ 3:9 Hz clearly point
to the practically free rotation of the base about the
glycosyl bond.17–20
1
The assignment of all the H and 13C resonances of c3A
There is a close stereochemical similarity between the
ribofuranose moieties of adenosine residues in the mol-
ecule of natural trimer, (20,50)A3,21 and analogues, con-
taining c3A in the middle position of the chain, A20p50
(c3A)20p50A, and at the 20-terminus, A20p50A20p50(c3A)
(Tables 2–5); data for adenosine residues of
(c3A)20p50A20 p50A are not available owing to the
is straightforward. The conformational analysis of the
furanose rings of compounds described was performed
with the aid of the PSEUROT (version 6.3) program,
3
which calculates the best fits of three JHH experimental
coupling constants [3J1 ;2 , J2 ;3 , and J3 ;4 ] to the five
conformational parameters (phase angles P and puck-
ering amplitudes wm for both the N- and the S-type
conformers and the corresponding molar ratio).15;16
Furanose ring of c3A is in the S $ N pseudorotational
equilibrium with a slight preference for the former [PS
3
3
0
0
0
0
0
0
1
overlap of the relevant H resonances. The sugar rings
of all nucleoside residues within the trimers appear to be
rather flexible (Table 5). The PSEUROT analyses of the
S $ N pseudorotational equilibrium of the furanose ring
of c3A fragment within A20p50A20p50(c3A) and the
122.3ꢁ, E1 (C10-exo) conformation]. The 3JC4;H1 value of
0

E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
3641
Table 2. Proton chemical shifts (dTMS, ppm) of 3-deazaadenosine (c3A) and analogues of (2-5)A3 containing c3A at different positions of the
oligonucleotide chain (D2O) (purine numbering)a
Compd
Residue
Chemical shifts
H-8
H-2
H-3
H-10
H-20
H-30
H-40
H-50
H-500
c3A
––
8.33 s
7.75 d
7.04 d
6.00 d
4.63 t
4.40 dd
4.47 m
3.94 dd
3.87 dd
20
A1
A2
c3A
8.27
8.14
7.75 s
7.99 s
7.32 d
––
––
6.11 d
5.97 d
5.84 d
5.07 dt
4.77 m
4.29 t
4.62 t
4.60 t
4.37 dd
4.20 m
4.18
4.28
3.73 dd
3.88 ddd
4.25 m
3.85 dd
4.09 ddd
4.16 ddd
8.00
(Dd ¼ 0:33)
8.11
7.00 d
(Dd ¼ 0:04)
––
(Dd ¼ 0:43)
7.65 s
21
22
A1
c3A
6.04 d
5.90 d
5.00 dt
4.44 ddd
4.55 t
4.33 t
4.20 m
4.07 m
3.68 dd
4.05 m
3.78 dd
3.90 m
7.99
(Dd ¼ 0:34)
8.08
7.22 d
(Dd ¼ 0:53)
8.09 s
6.88 d
(Dd ¼ 0:16)
––
A3
5.84 d
5.95 d
4.48 t
4.32 t
4.22 m
4.20 m
4.05 m
3.84 dd
4.05 m
3.71 dd
c3A
8.10
7.26 d
6.91 d
4.98 ddd
4.62 dd
(Dd ¼ 0:23)
8.00
7.91
(Dd ¼ 0:49)
7.74 s
8.26 s
(Dd ¼ 0:13)
A2
A3
––
––
5.98 d
5.74 d
4.58 ddd
4.26
4.28 m
4.15
4.20 m
4.10
n.d.
n.d.
n.d.
n.d.
a The chemical shifts of the b- -ribofuranose moieties of adenosine as well as c3A residues are in good agreement with those for the corresponding
D
adenosine residue of (20,50)A3;21 Dd ¼ dc3AðfreeÞ ꢀ dc3Aðwithin trimerÞ; n.d.––not determined.
Table 3. Selected coupling constants (J, Hz)a
arrangement of this molecule. Stereochemistry of the 3-
deazaadenosine residue within the trimer is very similar to
that of monomeric nucleoside c3A as well as of the rele-
vant adenosine residue of the natural trimer (20,50)A3
(Tables 4 and 5). However, some interesting features are
connected with this residue. Thus, a reversed order of the
H-20 and H-30, and C-20 and C-30 chemical shifts com-
pared to the parent monomeric c3A can be noticed. This
fact may be explained by the predominant syn confor-
mation of the c3A residue within the trimer. This sug-
Compd Residue
Coupling constants
10,20
5.80
20,30
30,40
4.29
3J31P;H
c3A
––
5.66
––
20
A1
A2
4.42
4.35
4.95
5.17
4.95
5.17
9.01 (P,H-20)
9.0 (P,H-20)
3
0
JP;H5 3.84
JP;H5 3.0
JP;H5 3.0
3
00
c3A
6.44
5.48
3.37
3
0
3
00
JP;H5 3.78
3
0
gestion is corroborated by the relationship JC8;H1
<
3
0
JC4;H1 (Table 4), which points to the predominant pop-
ulation of the syn base orientation about the glycosyl
bond.17–20 Moreover, the large upfield shift (3.6 ppm) of
the C-20 resonance of c3A residue of A20p50 A20p50(c3A)
versus the nucleoside c3A give further support to a pref-
erential syn orientation of the base (cf. Ref. 17). The
extensive 13C NMR studies by Uesugi and Ikehara of
various (bromo, methyl, chloro, thiomethyl, and meth-
oxy) 8-substituted purine nucleosides led to conclusion
that such an upfield shift of C-20 resonance compared to
those of the parent nucleosides is characteristic for a syn
conformation about the glycosyl bond.22 It is noteworthy
that conformational studies of different analogues of 2-
5A have led to the hypothesis that the syn base orientation
about the glycosyl bond at the 20-terminus resulted in an
enhancement of activating activity for RNase L.23 In the
present work we have found that p50A20p50A20p50(c3A)
retained activity equivalent to 2-5A itself (see below).
21
22
A1
c3A
5.07
2.94
5.10
5.11
4.66
6.02
9.06 (P,H-20)
8.03 (P,H-20)
3
0
JP;H5 and
JP;H5 ca. 3.0
3
00
b
A3
5.04
5.11
4.79
c3A
A2
A3
3.15
2.33
2.96
5.15
4.77
n.d.
6.00
n.d.
n.d.
8.71 (P,H-20)
7.09 (P,H-20)b
b
a The 3J2;3 values of c3A and c3A residue within the trimers 20–22 were
found to be 5.97, 7.04, 6.99, and 7.05 Hz, respectively.
b The JP;H5 and JP;H5 were not determined owing to the overlap of
3
3
0
00
the H-50 and H-500 resonances.
monomeric nucleoside, c3A, manifest close similarity.
On the contrary, the stereochemical behavior of the
furanose ring of c3A residue within (c3A)20p50A20p50A
and A20p50(c3A)20p50A trimers substantially differs from
that of the nucleoside itself as well as of adenosine,
thereby the pseudorotational parameters of c3A residue
within these trimers are practically identical. Of interest
is an observation of a very impressive upfield shift of the
H-2 and C-2 resonances of c3A residues of the trimers in
comparison with the monomeric c3A molecule (Tables 2
and 4). Such remarkable upfield shifts characterize
strong stacking interaction and manifest the profound
influence of neighboring adenine base onto this frag-
ment of c3A molecule.
The NMR data available clearly point to a close ste-
reochemical resemblance of adenosine residues of
A20p50(c3A)20p50A to the natural trimer (20,50)A3, but
rather essential deviations of the ribofuranose ring
conformation of c3A residue from that of the mono-
meric nucleoside c3A and the middle adenosine residue
as well (Tables 2–5) (cf. Ref. 21). The position of the
C-20 and C-30 resonances of c3A residue within
A20p50(c3A)20p50A molecule versus that of c3A itself
allow to suggest the anti base orientation about the
glycosyl bond of the middle nucleoside.
In the A20p50A20p50(c3A) trimer, the available NMR data
supply reliable information regarding the spatial

3642
E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
Table 4. Chemical shifts (dTMS, ppm) of 3-deazaadenosine (c3A) and analogues of (2-5)A3 containing c3A at different positions of the oligonucleotide
1
3
chain (D2O) (purine numbering) [some JC;H and JC;H values are given in brackets]
Compd Residue Chemical shift
C-6a
C-4
C-2
C-3
C-8 [1JC;H and
C-5b
C-10
C-20
C-30
C-40
C-50
3
3
2
3
3
2
0
JC8;H1
]
f JC;P
88.8
g
f JC;P
73.7
g
f JC;P
70.0
g
f JC;P
85.0
g
f JC;P
61.1
g
c3A
––
151.3 138.3c
140.3
[179.0]d
99.1
[171.3]e
141.0
[213.0 and 3.9]
126.6
[165.2]
[149.0]
[152.0]
[150.3]
[142.7]
20
A1f
A2f
c3A
155.0 147.9
154.5 147.8
148.1 138.3g
152.0
[202.8]
152.6
––
143.6
[214.0 and 2–3]
140.4
118.7
117.4
126.3
87.5
{6.3}
86.1
77.1
{4–5}
78.4
69.6
85.0
60.9
––
69.4
73.7
82.5
64.1
[203.0]
129.4
[212.4 and 4.0]
138.8
[214.6 and 4.6]
{8.3}
89.6
{4–5}
70.1
{9.4}
84.0
{9.7}
{4–5}
64.5
{4–5}
99.6
(Dd ¼ 10:9)
152.0
21
22
A1f
c3A
A3
155.2 147.8h
148.5 137.8
155.0 147.6h
––
142.7
139.1
141.1
118.8
125.6
118.0
87.6
{4.9}
88.9
77.1
{5.0}
77.5
70.3
85.5
61.2
129.6
(Dd ¼ 10:7)
152.8
99.3
––
67.7
{2–3}
69.8
81.7
63.0
{6.7}
87.1
{5.3}
74.1
{9.6}
83.0
{8.9}
{4.3}
64.6
{4.8}
c3A
A2f
A3
147.9 138.4
155.2 147.9h
154.2 147.7h
129.5
(Dd ¼ 10:8)
152.7
99.2
––
138.6
139.2
142.2
123.6
118.2
117.2
88.0
{3–4}
86.2
77.8h
{5.6}
77.5h
{5.1}
74.8
69.1
{2–3}
68.5
83.7
59.8
82.5
64.0
{5–6}
87.6
{5.0}
69.2
{9.2}
81.7
{9.4}
{4–5}
63.7
{2–3}
152.3
––
a
3
The JC6;H2 value was found to be ca. 11 and 9 Hz in the case of adenine and c3A residues, respectively.
b
3
3
3
The JC5;H8 and JC5;H3 values of c3A were found to be 11–12 and 5–6 Hz, respectively; the JC5;H8 value of adenosine was ca. 11 Hz.
c The C-4 resonance appeared as a multiplet with the following 13C–1H couplings: 9.7 Hz (3JC4;H8), 2.5–3.7 Hz (3JC4;H1 ), and 5.6–6.6 Hz ( JC4;H2); note
3
0
that the JC4;H1 value along with the JC8;H1 ¼ 3:9 Hz clearly point to the practically free rotation of the base about the glycosyl bond.17–20 The
3
3
0
0
2JC4;H3 is less than 2.0 Hz.
d
e
2
The JC2;H3 value is less than 2.0 Hz. An Dd is a difference between the C-2 chemical shift of c3A and c3A residue within the trimer.
2
The JC3;H2 was found to be 7.8 Hz in the case of c3A and 4.1 Hz for c3A residue of ApAp(c3A).
f The chemical shifts and coupling constants are in fair agreement with those for the corresponding adenosine residue of (20,50)ApAp[8-(4-amino-
butyl)amino]adenosine.17
g The C-4 resonance appeared as a multiplet with the following 13C–1H couplings: 10.9 Hz (3JC4;H8) and ca. 6.0 Hz (3JC4;H1 and 3JC4;H2); note that the
0
3
3
0
0
JC4;H1 of ca. 6 Hz along with the JC8;H1 ¼ 4:6 Hz imply the predominant syn base orientation about the glycosidic bond.
h The data may be interconvertible.
Table 5. Pseudorotational parameters of the b-D
-ribofuranose moieties of 3-deazaadenosine (c3A) and analogues of (2-5)A3 containing c3A at
different positions of the oligonucleotide chain in D2O solutionsa
Compd
c3A
Residue
––
PN
wmðNÞ
PS
wmðSÞ
% S
)34.3 (1T2)
36
122.3 (E1)
40
60
20
21
22
A1b
A2b
c3Ab
)18.1 (E2)
)21.8 (E2)
)39.5 (1T2)
38
38
36
137.0 (1T2)
127.1 (E1)
130.9 (E1)
38
38
43
43
43
65
A1b
c3Ac
A3b
)19.4 (E2)
35.3 (3T4)
)18.0 (E2)
38
38
38
133.6 (1T2)
214.8 (4T3)
132.3 (1T2)
38
38
38
51
38
51
c3Ac
A2
38.2 (3T4)
n.a.
n.a.
38
212.7 (4T3)
38
38
A3
a The conformational analysis of the furanose rings of compounds studied was performed by the PSEUROT (version 6.3) program. The underlined
values were fixed during the final calculations; the rms deviations and jDJmaxj data were found to be 0.00 for all calculated couplings. n.a.––not
analyzed owing to the impossibility to measure the necessary H–H coupling constants.
b Pseudorotational parameters are in fair agreement with those for the corresponding adenosine residue of (20,50)A3.21
c Pseudorotational parameters display rather essential deviations from those of both the parent c3A and the corresponding adenosine residue of
(20,50)A3.21
The NMR spectral data for (c3A)20p50A20p50A trimer
are rather scanty, which precludes an elucidation of the
spatial arrangement of this molecule. Nonetheless, the
anti base orientation about the glycosyl bond of c3A

E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
3643
40
30
20
10
0
ApAp(c3A)
Ap(c3A)pA
(c3A)pApA
200
220
240
260
280
300
-10
-20
-30
-40
Wavelength [nm]
Figure 2. RNase L activating ability of analogues of 2-5A. 2-5A and
analogues at different concentrations (as indicated) were incubated
with human recombinant RNase L on ice for 30 min followed by
incubation with C7UUC12-[32P]pCp at 30 ꢁC for 30 min. An autora-
diogram of the dried gel is shown (from left to right): lane 1––control
with no activator present; lanes 2–6, 1000–0.1 nM (20,50)pA4, respec-
tively; lanes 7–10, 1000–1 nM p(c3A)pApA, respectively; lanes 11–14,
1000–1 nM pAp(c3A)pA, respectively; lanes 15–18, 1000–1 nM pA-
pAp(c3A), respectively. Arrows indicate the intact and cleaved RNA,
Figure 1. CD spectra of 3-deazaadenosine-substituted analogues of the
core trimer A20p50A20p50A in water at 20 ꢁC.
residue can be suggested taking into account the C-20
and C30 resonances. One common characteristic of tri-
mers 20–21 is a similarity of stereochemistry of the sugar
3
0
phosphate backbones that can be seen from the JC1 ;P
,
3
3
3JC3 ;P, JC4 ;P, and JH2 ;P values as well.
C12[
32P]pCp. The graph represents percent RNA cleavage as deter-
mined by phosphorimager analysis.
0
0
0
The CD spectra of trimers 20 and 22 are similar in shape
to that of the natural trimer, (20,50)A3,24;25 albeit display
essential differences in amplitudes of the Cotton effects
(Fig. 1). Unexpectedly, the CD spectrum of
A20p50(c3A)20p50A trimer containing c3A in the middle
position of the chain shows a mirror-like shape and the
long-wave Cotton effect is missing at all. Such CD
behavior of the latter points to essential differences in its
spatial arrangement compared to two other trimers,
which is in agreement with the biochemical data (see
below). It is noteworthy that the hypochromism data
display remarkable diversity depending on the point of
modification of (20,50)A3 (Table 1), thereby A20p50(c3A)-
20p50A trimer occupies the mid-position. It should,
however, be stressed that our previous attempts to find
out the possible correlations between the NMR, CD,
and hypochromicity have failed (cf. Ref. 24).
According to the results of Figure 2, parent 2-5A tet-
ramer, p50A20p50A20p50A20p50A, was able to effect a 50%
cleavage of pC7UUC12pCp at a concentration of 1 nM.
It may be noted that multiple earlier studies have shown
that both 2-5A trimer, p50A20p50A20p50A and
p50A20p50A20p50A20p50A, are close in bioactivity. Simi-
larly, the trinucleotides, p50(c3A)20p50A20p50A and
p50A20p50A20p50(c3A) were of quite similar potency to
parent 2-5A tetramer, effecting substrate cleavages of
60% and 50% at 1 nM, respectively. In distinct contracts,
the trimeric congener, p50A20p50(c3A)20p50A caused only
20% cleavage at 1 nM and 80% cleavage at 10 nM. Thus,
this analogue, in which the middle adenosine had been
altered to c3A, was at least 10-fold less effective than
either the 50- or 20-terminally modified analogues,
thereby testifying to a special influence of this middle 3-
deazaadenosine nucleoside on the bioactivity of 2-5A.
2.3. Biological studies
Although mouse RNase L was first purified to homo-
geneity,26 it was the human enzyme, after expression of
the cloned enzyme in SF21 cells infected with re-
combinant baculovirus27 that was first prepared in suf-
ficient quantity for meaningful enzymological studies.
Enzyme used in this study was purified according to
such published methods.
3. Conclusions
Sequence-specific 3-deazaadenosine-substituted ana-
logues of trimeric 20,50-oligoadenylate, p50A20p50A20
p50A, and the respective 50-dephosphorylated, core tri-
mers were synthesized employing the phosphotriester
methodology.
The extensive conformational analysis of the c3A-
substituted core trimers versus the parent natural trimer
by the 1H and 13C NMR spectroscopy and CD displayed
close stereochemical similarity between the natural core
trimer and (c3A)20p50A20p50A and A20p50A20p50(c3A)
analogues. The NMR data for the latter core trimer,
A20p50A20p50(c3A), clearly point to the syn base orien-
tation about the glycosyl bond of the c3A residue. This
observation is in a fair agreement with the hypothesis23
that the syn base orientation of the A3 residue of 2-5A
Methods of assay of RNase L activity have relied upon
cleavage of radiolabeled poly(U)26 or on cleavage of
ribosomal RNA (see Ref. 28). A method that would
utilize a synthetic RNA of defined length in which
cleavages would yield products of discrete lengths has
been developed by Carroll et al.29 Since it was demon-
strated earlier that RNase L cleaves preferentially after
UU and UA sequences, the sequence chosen for analysis
of RNase L was 50-[32P]pC11UUC7. Detailed procedures
have been published for these assays.27

3644
E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
contributes to the activation of RNase L, on the one
hand, and with the finding that p50A20p50A20p50(c3A)
was nearly as active as the natural tetramer, on the
other. An analogue A20p50(c3A)20p50A displayed rather
essential deviations from the spatial arrangement of the
natural core trimer.
prepared according to Mizuno et al.30 Yield 34%; mp
225–226 ꢁC (EtOH); TLC (E): Rf 0.71; UV (H2O), kmax
,
nm (e ꢁ 10ꢀ3): (pH 1.0 and 7.0) 262 (10.8), (pH 13.0) 265
(10.7); CD (K/Na-phosphate buffer, pH 7.4), kmax, nm
(H ꢁ 10ꢀ3): 218 (0), 220 ()6.97), 230–245 (0), 263
()2.66), 290 (0).
Substitution of either 50-terminal or 20-terminal adeno-
sine with c3A afforded the respective analogues
p50(c3A)20p50A20p50A and p50A20p50A20p50(c3A) that
were as effective as the natural tetramer itself as acti-
4.1.3. 4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-(b-
D
-ribofuranosyl)-1H-imidazo[4,5-c]pyridine (5) (cf. Ref.
8). Trimethylsilyl chloride (0.9 mL, 7.1 mmol) was added
to a solution of 4 (0.25 g, 0.94 mmol) in anhyd pyridine
(2.5 mL) and the reaction mixture was stirred for 1.5 h.
Then, 2-(4-nitrophenyl)ethylchloroformate (1.03 g,
4.7 mmol) was added and stirring was continued over-
night. The reaction mixture was cooled to 0 ꢁC and cold
water (1.64 mL) was added under stirring. After 10 min,
concd aq ammonia (3.16 mL) was added and stirring
was continued for 30 min. The reaction mixture was
evaporated and the residue was purified by silica gel
column chromatography (80 mL). Elution was per-
formed with a linear MeOH gradient (0–50%, v/v, 1 L)
in CHCl3. The fractions containing the product were
collected, evaporated, and crystallized from EtOH to
give 5 as colorless crystals (0.296 g, 68%); mp 112–
114 ꢁC (EtOH); TLC (B): Rf 0.23; UV (H2O), kmax, nm
(e ꢁ 10ꢀ3): 263 (9.0); kmin 232 (1.96); (EtOH), kmax
271 nm; kmin 232 nm. 1H NMR (DMSO-d6): 8.56 (s, 1H,
H–C(8)), 8.08 (d, 1H, J2;3 ¼ 5:5, H–C(2)), 7.66 (d, 1H,
H–C(3)), 8.16 (d, 2H, J ¼ 8:5; ortho-Ph–NO2), 7.60 (d,
vators of RNase
L
(EC50 ¼ 1 nM). In contrast,
p50A20p50(c3A)20p50A showed diminished RNase L
activation ability (EC50 ¼ 10 nM). This finding is in har-
mony with essential stereochemical differences between
A20p50(c3A)20p50A and the natural core (20,50)trimer,
whereas specific recognition of the nitrogen-3 atom of
the middle adenosine A2 seems to be unlikely.
4. Experimental
4.1. Synthesis of 3-deazaadenosine oligonucleotides 20–22
and their 50-monophosphates 1–3
4.1.1. General. The UV spectra were recorded on a
Specord UV–vis spectrophotometer (Carl Zeiss, Ger-
many). 1H and 13C NMR spectra were measured at
200.13 MHz on an AC 200 (blocked derivatives of c3A
5–11) and at 500.13 and 125.76 MHz on an Avance 500
DRX (trimers 20–22) spectrometers (Bruker, Germany)
with tetramethylsilane as an internal standard (s ¼ sin-
glet, d ¼ doublet, t ¼ triplet, q ¼ quartet, m ¼ multiplet,
br s ¼ broad signal); the chemical shifts (d) and coupling
constants (J) are given in ppm rel. to external TMS and
Hz, respectively. Assignments of proton resonances
were confirmed, when possible, by selective homonu-
clear decoupling experiments and [1H,1H] correlation
spectra as well; assignments of proton and 13-carbon
resonances have been proved by [1H,13C] correlation
spectra. The CD spectra were recorded on Jasco J-20
spectropolarimeter (Jasco, Japan) in water solutions.
Thin layer chromatography (TLC) was carried out on
the 60F254 silica gel plates (Merck, Germany). As solvent
systems were used: CHCl3/MeOH, 9:1 (A), CHCl3/
MeOH, 4:1 (B), EtOAc/hexane, 4:1 (C), CHCl3/MeOH/
Et3N, 9:0.3:0.3 (D), i-PrOH/NH3/H2O, 12:1:2 (E). Col-
umn chromatography was performed on silica gel
Merck 60 (0.040–0.063 mm). Melting points were
determined with a Boethius (Germany) apparatus and
are uncorrected. The solutions of compounds in organic
solvents were dried with anhydrous sodium sulfate for
4 h. The reactions were performed at room temperature,
unless stated otherwise. High-performance liquid chro-
matography (HPLC) was carried out with Waters
apparatus with a column Nova-Pac C-18 (3.9 · 300 mm)
using an isocratic gradient elution with the following
buffer: 7% CH3CN in 0.05 M KH2PO4, v/v at a flow rate
0.7 mL/min.
0
0
2H, J ¼ 8:5; meta-Ph–NO2), 5.90 (d, 1H, J1 ;2 ¼ 6:5; H–
C(10)), 5.56 (d, 1H, J2 ;2 -OH ¼ 4:5; HO–C(20)), 5.31 (br
3
0
0
d, 1H, J3 ;3 -OH ¼ 2:0; HO–C(30)), 5.20 (br t, 1H,
3
0
0
J5 ;5 -OH ¼ 2:5; HO–C(50)), ꢂ4.36 (m, 1H, H–C(20)),
3
0
0
ꢂ4.36 [br t, 3H (+H-20), J ¼ 4:0; –CH2–CH2–PhNO2],
4.12 (br m, 1H, J3 ;4 ¼ 1:5; H–C(30)), 4.00 (br m, 1H, H–
C(40)), 3.67 (br m, 2H, H2–C(50), 3.10 (t, 2H, J ¼ 4:0;
–CH2–CH2–PhNO2).
0
0
Anal. Calcd for C20H21N5O8 (459.41): C, 52.29; H, 4.61.
Found: C, 52.02; H, 4.62.
4.1.4. 4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-[5-
O-(4-monomethoxytrityl]-(b- -ribofuranosyl)-1H-imidazo-
D
[4,5-c]pyridine (6). A solution of compound 5 (70 mg,
0.15 mmol) and 4-monomethoxytrityl chloride (80 mg,
0.25 mmol) in anhyd pyridine (0.5 mL) was stirred
for 24 h. The reaction mixture was evaporated,
co-evaporated with MeOH (2 · 10 mL), and the residue
was purified by silica gel column chromatography
(30 mL). Elution was performed with a linear MeOH
gradient (0–10%, v/v, 0.5 L) in CHCl3. The fractions
containing the product were collected, evaporated. The
residue was dissolved in EtOAc (1 mL) and precipitated
in hexane (50 mL). The resulting precipitate was col-
lected by filtration and dried in vacuo to give 6 as col-
orless crystals (80 mg, 73%); mp 136–138 ꢁC; TLC (A):
1
Rf 0.33. H NMR (CDCl3): 8.16 (s, 1H, H–C(8)), 8.0–
7.16 (m, 18H, 2·C6H5 the ortho- and meta-Ph–NO2, the
ortho protons of MTr group, H–C(2) and H–C(3)), 6.72
(d, 2H, 3Jmeta;ortho ¼ 8:5; the meta protons of MTr group),
4.1.2.
4-Amino-1-(b- -ribofuranosyl)-1H-imidazo[4,5-
D
c]pyridine [c3A; (4)]. Compound [c3A; (4)] was
5.80 (d, 1H, J1 ;2 ¼ 6:0; H–C(10)), 4.59 (dd, 1H,
0
0

E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
3645
J2 ;3 ¼ 3:0; H–C(20)), ꢂ4.50 (br t, 2H, J ꢂ 6:0; –CH2–
5.94 (d, 1H, J1 ;2 ¼ 6:0; H–C(10)), 4.98 (br t, 1H,
0
0
0
0
CH2–PhNO2), 4.52 (dd, 1H, J3 ;4 ¼ 2:5; H–C(30)), 4.31
J2 ;3 ¼ 4:0; H–C(20)), 5.72 (br t, 1H, J3 ;4 ¼ 1:5; H–
0
0
0
0
0
0
3
3
(m, 1H, J4 ;5 ¼ J4 ;5 ¼ 1:5; H–C(40)), 3.75 (s, 3H,
OCH3), 3.48 (d, 2H, H2–C(50)), 3.10 (t, 2H, J ¼ 6:0;
–CH2–CH2–PhNO2).
C(30)), 4.51 (br m, 1H, J4 ;5 ¼ J4 ;5 ¼ 2:0; H–C(40)),
0
0
0
00
0
0
0
00
4.38 (br t, 2H, J ¼ 4:25; –CH2–CH2–PhNO2), 3.74 (s,
gem
0
00
3H, OCH3), ꢂ3.56 (center of m, 2H,
J
¼ 10:0; H–
5 ;5
C(50)), 3.04 (t, 2H, J ¼ 4:25; –CH2–CH2–PhNO2).
Anal. Calcd for C40H37N5O9 (731.75): C, 65.65; H, 5.10.
Found: C, 65.92; H, 5.42.
Anal. Calcd for C47H41N5O10 (835.86): C, 67.54; H,
4.94. Found: C, 67.91; H, 4.62.
4.1.5. Benzoylation of 4-{[2-(4-nitrophenyl)ethoxycarb-
onyl]amino}-N1-[5-O-(4-monomethoxytrityl]-(b-
D-ribofur-
4.1.8.
[(2,3-di-O-benzoyl)-(b-
pyridine 9. Compound 9 was prepared by standard
detritylation12 of 7 (94 mg, 0.1 mmol) and isolated as an
amorphous powder; yield 60 mg (90%); mp 103–104 ꢁC;
TLC (A): Rf 0.45. 1H NMR (DMSO-d6): 9.74 (br s, 1H,
4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-
-ribofuranosyl)]-1H-imidazo[4,5-c]-
anosyl)-1H-imidazo[4,5-c]pyridine (6) with benzoyl chlo-
ride.8;9;17 To the stirred solution of compound 6 (116 mg,
0.16 mmol) in a mixture of anhyd CH3CN (2.1 mL),
Et3N (0.28 mL), and 4-dimethylaminopyridine (DMAP)
(1.4 mg), freshly distilled BzCl (0.02 mL, 24 mg,
0.17 mmol) was added. After stirring for 1 h, the reac-
tion mixture was evaporated, co-evaporated with
MeOH (2 · 20 mL), and the products were purified by
silica gel column chromatography (50 mL). Elution was
performed with a linear EtOAc gradient (20–80%, v/v,
1 L) in hexane. In order of elution were isolated:
D
H–N(6)), 8.68 (s, 1H, H–C(8)), 8.12 (m, 3H, ortho-Ph–
3
NO2 and H–C(2)), 8.04 and 7.80 (2dd, 4H, Jortho;meta
¼
3
7:5, Jpara;ortho ꢂ 1:0; the ortho protons of Bz groups),
7.76–7.32 (m, 9H, the meta and para protons of Bz
groups, H–C(3) and meta-Ph–NO2), 6.64 (d, 1H,
J1 ;2 ¼ 7:0; H–C(10)), 5.96 (dd, 1H, J2 ;3 ¼ 5:0; H–C(20)),
0
0
0
0
5.84 (dd, 1H, J3 ;4 ¼ 1:5; H–C(30)), 5.70 (t, 1H, J5 ;5 -OH
¼
0
0
0
0
2:5; HO–C(50)), 4.56 (br m, 1H, J4 ;5 ¼ J4 ;5 ¼ 2:0; H–
C(40)), 4.34 (br t, 2H, J ¼ 6:25; –CH2–CH2–PhNO2),
3.88 (center of m, 2H, H2–C(50)), 3.06 (t, 2H, J ¼ 6:25;
–CH2–CH2–PhNO2).
0
0
0
00
4.1.6. 4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-
[(2,3-di-O-benzoyl)-5-O-(4-monomethoxytrityl]-(b-D-ribo-
furanosyl)-1H-imidazo[4,5-c]pyridine (7). Compound 7
was dissolved, after evaporation of combined fractions,
in EtOAc (1 mL) and precipitated from hexane (50 mL)
to give a solid (20 mg, 13%); mp 110–112 ꢁC; TLC (C):
Rf 0.29; UV (EtOH), kmax 271 and 231; kmin 251 and 225;
1H NMR (CDCl3): 8.16 (s, 1H, H–C(8)), 8.18 [d, 3H,
(ortho-Ph–NO2 and H-8), J ꢂ 7:5], 8.04 and 7.92 (2d,
5H, J ꢂ 7:5; the ortho protons of Bz groups and H-2),
ꢂ7.60 (d, 1H, J2;3 ꢂ 6:0; H–C(3)), 7.54–7.16 (m, 21H,
2·C6H5, meta-Ph–NO2 and the ortho protons of Bz
Anal. Calcd for C34H29N5O10 (667.62): C, 61.17; H,
4.38. Found: C, 61.22; H, 4.63.
4.1.9. 4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-
{2-O-[2-(4-nitrophenylethyl)-phosphato]-[3-O-benzoyl-5-O-
(4-monomethoxytrityl)-b-
[4,5-c]pyridine (11). Compound 11 was prepared as de-
scribed earlier.9–12 Phosphorylation of
(0.11 g,
0.132 mmol) with 2-chlorophenyldi(triazolido)phos-
phate followed by the treatment with 2-(4-nitrophe-
nyl)ethanol afforded the triester 10 as oil (0.112 g, 73%);
TLC (C): Rf 0.47. The latter (0.13 g, 0.11 mmol) was
treated with p-nitrobenzaldoxime in a mixture of tri-
ethylamine/pyridine/water (1:1:1, vol) followed by work-
up to give the diester 11 (triethylammonium salt) as an
D
-ribofuranosyl]-1H-imidazo-
8
3
groups), 6.82 (d, 2H, Jmeta;ortho ¼ 8:5; the meta protons
of MTr group, the meta and para protons of Bz groups),
6.29 (d, 1H, J1 ;2 ¼ 6:0; H–C(10)), 6.19 (br t, 1H,
0
0
J2 ;3 ¼ 6:0; H–C(20)), 6.03 (br t, 1H, J3 ;4 ꢂ 3:0; H–
C(30)), 4.50 (br t, 2H, J ¼ 6:25; –CH2–CH2–PhNO2),
4.60 (br m, 1H, H–C(40)), 3.77 (s, 3H, OCH3), ꢂ3.66
(center of m, 2H, H–C(50)), 3.13 (t, 2H, J ¼ 6:25; –CH2–
CH2–PhNO2).
0
0
0
0
1
amorphous powder (91 mg, 71%). TLC (D): Rf 0.27. H
NMR (DMSO-d6): 9.78 (br s, 1H, H–N(6)), 8.56 (s, 1H,
H–C(8)), 8.10–7.18 (m, arom. H), 6.78 (d, 2H,
Jmeta;ortho ¼ 7:5; the meta protons of MTr group), 6.38 (d,
Anal. Calcd for C54H45N5O11 (939.96): C, 69.00; H,
4.83. Found: C, 69.22; H, 5.12.
1H, J1 ;2 ¼ 7:5; H–C(10)), 5.72 (br s, 1H, H–C(20)), 5.40
(br s, 1H, H–C(30)), 4.40 (br s, 1H, H–C(40)), 4.36 (br t,
2H, J ¼ 6:0; –CH2–CH2–PhNO2), 3.06 (t, 2H, J ¼ 6:0; –
CH2–CH2–PhNO2); resonances of the second
–CH2–CH2–PhNO2 group are overlapped by an intense
resonance of OH group and DMSO-d5; resonances of
H2–C(50) are overlapped by an intense resonance of OH
group.
0
0
4.1.7. 4-{[2-(4-Nitrophenyl)ethoxycarbonyl]amino}-N1-
[(3-O-benzoyl)-5-O-(4-monomethoxytrityl]-(b-D-ribofuran-
osyl)-1H-imidazo[4,5-c]pyridine (8). Compound 8 was
dissolved, after evaporation of combined fractions, in
EtOAc (1 mL) and precipitated from hexane (50 mL) to
give a solid (80 mg, 60%); mp 110–112 ꢁC; TLC (C): Rf
0.14; UV (EtOH), kmax 270; kmin 249. 1H NMR (CDCl3):
8.20–8.00 (m, 6H, ortho-Ph–NO2, the ortho protons of
Bz group, H–C(8) and H–C(2)), 7.73 (d, 1H, J ¼ 5:5; H–
C(3)), 7.60–7.14 (m, 17H, 2·C6H5, meta-Ph–NO2, the
ortho protons of MTr group and Bz group, the meta-
Ph–NO2, the meta and para protons of Bz group), 6.77
(d, 2H, 3Jmeta;ortho ¼ 8:5; the meta protons of MTr group),
4.1.10. N6,N6,O20,O30-Tetrabenzoyladenosine (12) and
N6,O30-dibenzoyl-50-O-(4-monomethoxytrityl)-20-O-[2-(4-
nitrophenylethyl)-phosphato]adenosine
nium salt) (13). Compounds 12 and 13 have been pre-
(triethylammo-
pared as described previously.9;10;17

3646
E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
4.1.11. Synthesis of dimers 14–16, trimers 17–19 and
deprotected core (20,50)trimers 20–22. Synthesis of dimers
(14–16) and then trimers (17–19) was performed
according to the previously described methodology7–10;17
consisting of (i) condensation of either phosphodiester
11 with 20-terminal building block 12, or 13 with 9, or 12
with 13 and subsequent detritylation, and (ii) conden-
sation of individual dimer obtained with phosphodiest-
ers 11 or 13, and subsequent detritylation.
and stirred for 18 h. After the addition of 1 M solution
of acetic acid in pyridine (12 mL), the mixture was
evaporated and then co-evaporated with pyridine
(2 · 5 mL). The residue was dissolved in aq 25% solution
of ammonia, kept for 18 h, evaporated, and purified by
column chromatography as described above for core
trimers 20–22. Deblocked 50-monophosphate (2) was
obtained in form of Naþ-salt as an amorphous powder.
To a solution of the appropriate phosphodiester
(0.11 mmol) and the 50-terminal building block or dimer
(0.1 mmol) in CHCl3 (0.5 mL), 2,4,6-triisopropyl-
benzenesulfonyl chloride (TPS-Cl, 0.3 mmol) and N-
methylimidazole (0.9 mmol) were added and the reac-
tion mixture was stirred for 20 min. The reaction mix-
ture was poured into hexane (200 mL), the resulting
precipitate was collected by filtration, dried in vacuo,
and then dissolved in 2% solution of p-toluenesulfonic
acid in CH2Cl2/MeOH (7:3, v/v, 15 mL). After stirring
for 10 min (in the case of 14), 20 min (15), 30 min (16),
and 55 min (in the case of trimers 17–19), the solution
was diluted with CHCl3 (15 mL) and washed with
0.05 M phosphate buffer, pH 7.0 (2 · 30 mL). The or-
ganic layer was separated, dried, evaporated, and puri-
fied by silica gel column chromatography (60 mL). The
product was eluted with a linear methanol gradient (0–
5%, v/v, 2 · 300 mL) in CHCl3. Appropriate fractions
were collected, evaporated to a volume of 2 mL and
precipitated into hexane (200 mL). The compounds were
obtained as amorphous solids.
4.1.13. 50-O-Phosphoryladenylyl(20 fi 50)adenylyl(20 fi 50)-
3-deazaadenosine, sodium salt (3). In a similar way,
starting from the trimer 17 (30 mg, 0.015 mmol) the
compound 3 was obtained as an amorphous powder.
4.1.14. 50-O-Phosphoryl-3-deazaadenylyl(20 fi 50)adenylyl-
(20 ! 50)adenosine, sodium salt (1). In a similar way,
starting from the trimer 19 (30 mg, 0.014 mmol) the
compound 1 was obtained as an amorphous powder.
The isolated yields and some physico-chemical proper-
ties are given in Table 1.
4.2. Biochemical studies
4.2.1. General . Pure recombinant human RNase L was
prepared by a modification28 of a previously described
procedure.27 The 50-[32P]pCp (specific activity 3000 Ci/
mmol) was purchased from DuPont/NEN (Wilmington,
DE, USA). The synthetic oligonucleotide 50-
[32P]pC11UUC7 was prepared according to Carroll
et al.29
Deprotection of trimers 17–19 was performed by the
sequence deblocking of phosphate group and sub-
sequent treatment with methanol, saturated with
ammonia at 0 ꢁC. A trimer (0.1 mmol) was dissolved in
0.5 M solution of 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU) in pyridine (30 mL) and stirred for 2 h. After the
addition of 1 M solution of acetic acid in pyridine
(15 mL), the mixture was evaporated and then co-
evaporated with pyridine (2 · 10 mL). The residue was
dissolved in saturated methanolic ammonia (20 mL),
4.2.2. RNase L activation assays of 3-deazaadenosine
analogues 1–3 of 2-5A. Assay for activation of RNase L
was done using 32P-labeled C7UUC12 as an RNA sub-
strate.29 The synthetic RNA, C7UUC12, which was
prepared on an ABI model 380 DNA synthesizer, was
labeled at its 30-terminus with [50-32P]pCp (3000 Ci/
mmol) with T4 RNA ligase. Different 2-5A analogues at
1–1000 nM, compared with 0.1–1000 nM of pA4
[pA(20p50A)3] as a positive control, were incubated on
ice for 30 min with 0.2 mg/reaction of RNase L ex-
pressed from a baculovirus vector in insect cells.27;28;32
Reaction mixtures were further incubated with the RNA
substrate, 80 nM of C7UUC12[32P]pCp, for 30 min at
30 ꢁC. RNA was analyzed in 20% acrylamide/7 M urea
gels and the extent of degradation was measured with a
phosphorimager (Amersham Biosciences).
kept for 20 h and evaporated. The residue was chro-
ꢀ
matographed on a DEAE-Sephadex A-25 (HCO3
-
form, 100 mL) column using a linear gradient (0–0.8 M,
2 · 500 mL) of TEAB buffer. The product containing
fractions were collected and evaporated. Deblocked core
(20,50)trimers (20–22) were obtained in the form of Naþ-
salts as amorphous powders according to Moffatt.31 The
isolated yields and some physico-chemical properties are
given in Table 1.
4.1.12. 50-O-Phosphoryladenylyl(20 fi 50)-3-deazaadenyl-
yl(20 fi 50)adenosine, sodium salt (2). To a solution of
the trimer 18 (30 mg, 0.0137 mmol) in anhyd EtOAc
(0.3 mL) pyrophosphoryl chloride (0.247 mmol, 63 mg,
0.034 mL) was added and the reaction mixture was
stirred for 3 h at 0 ꢁC. The reaction was stopped by the
addition of ice, neutralized with cold 0.5 M TEAB
(4 mL), evaporated and co-evaporated with MeOH
(2 · 30 mL) and pyridine (2 · 10 mL). The residue was
dissolved in 0.5 M solution of DBU in pyridine (20 mL)
Acknowledgements
I. A. Mikhailopulo is deeply grateful to the Alexander
von Humboldt-Foundation (Bonn-Bad-Godesberg,
Germany) for a research fellowship.

E. N. Kalinichenko et al. / Bioorg. Med. Chem. 12 (2004) 3637–3647
3647
17. Mikhailopulo, I. A.; Kalinichenko, E. N.; Podkopaeva, T.
L.; Kelve, M.; Saarma, M.; Lidaks, M. Y.; Rosemeyer, H.;
Seela, F. Nucleosides Nucleotides 1997, 16, 1855.
18. Akhrem, A. A.; Mikhailopulo, I. A.; Abramov, A. F. Org.
Magn. Res. 1979, 12, 247.
References and notes
1. Williams, B. R. G.; Kerr, I. M. Nature 1978, 276, 88.
2. Player, M. R.; Torrence, P. F. Pharmacol. Ther. 1998, 78, 55.
3. Jamoulle, J. C.; Imai, J.; Lesiak, K.; Torrence, P. F.
Biochemistry 1984, 23, 3063.
4. Jamoulle, J. C.; Lesiak, K.; Torrence, P. F. Biochemistry
1987, 26, 376.
19. Davies, D. B. Progr. Nucl. Magn. Res. Spectrosc. 1978, 12,
135.
€
€ €
20. Tennila, T.; Azhayeva, E.; Vepsalainen, J.; Laatikainen,
R.; Azhayev, A.; Mikhailopulo, I. A. Nucleosides
Nucleotides 2000, 19, 1861.
5. Torrence, P. F.; Imai, J.; Lesiak, K.; Jamoulle, J. C.;
Sawai, H. J. Med. Chem. 1984, 27, 726.
21. van den Boogaart, J. E.; Kalinichenko, E. N.; Podkopa-
eva, T. L.; Mikhailopulo, I. A.; Altona, C. Eur. J.
Biochem. 1994, 221, 759.
22. Uesugi, S.; Ikehara, M. J. Am. Chem. Soc. 1977, 99,
3250.
6. Imai, J.; Lesiak, K.; Torrence, P. F. J. Biol. Chem. 1985,
260, 1390.
7. Player, M. R.; Kalinichenko, E. N.; Podkopaeva, T. L.;
Mikhailopulo, I. A.; Seela, F.; Torrence, P. F. Biochem.
Biophys. Res. Commun. 1998, 245, 430.
23. Van den Hogen, Y. T.; Hilgersom, C. M.; Brozda, D.;
Lesiak, K.; Torrence, P. F.; Altona, C. Eur. J. Biochem.
1989, 182, 629.
8. Mikhailopulo, I. A.; Kalinichenko, E. N.; Podkopaeva, T.
L.; Wenzel, T.; Rosemeyer, H.; Seela, F. Nucleosides
Nucleotides 1996, 15, 445.
24. Kalinichenko, E. N.; Podkopaeva, T. L.; Mikhailopulo, I.
A.; Kelve, M.; Saarma, M.; van den Boogaart, J. E.;
Altona, C. Russian J. Bioorg. Chem. 1999, 25, 675.
25. Kitade, Y.; Wakana, M.; Tsuboi, T.; Yatome, C.; Bayly,
S. F.; Player, M. R.; Torrence, P. F. Bioorg. Med. Chem.
Lett. 2000, 10, 329.
26. Silverman, R. H.; Jung, D. H.; Nolan-Sorden, N. L.;
Dieffenbach, C. W.; Kedar, V. P.; Sen Gupta, D. J. Biol.
Chem. 1988, 263, 7336.
27. Dong, A.; Xu, L.; Zhou, A.; Hassel, B. A.; Lee, X.;
Torrence, P. F.; Silverman, R. H. J. Biol. Chem. 1994, 269,
14153.
28. Player, M. R.; Bayly, S.; Wondrak, E.; Torrence, P. F.
Methods: Companion Methods Enzymol. 1998, 15, 243–
253.
29. Carroll, S. S.; Chen, E.; Viscount, T.; Geib, J.; Sardana,
M. K.; Gehman, J.; Kuo, L. C. J. Biol. Chem. 1994, 271,
4988.
30. Mizuno, Y.; Tazawa, S.; Kageura, K. Chem. Pharm. Bull.
1968, 16, 2011.
31. Moffatt, J. G. Can. J. Chem. 1964, 42, 599.
32. Dong, B.; Silverman, R. H. J. Biol. Chem. 1997, 272,
22236.
9. Kalinichenko, E. N.; Podkopaeva, T. L.; Poopeiko, N. E.;
Kelve, M.; Saarma, M.; Mikhailopulo, I. A.; van den
Boogaart, J. E.; Altona, C. Recueil Trav. Chim. Pays-Bas
1995, 114, 43.
10. Kvasyuk, E. I.; Kulak, T. I.; Khripach, N. B.; Mikhai-
lopoulo, I. A.; Uhlmann, E.; Charubala, R.; Pfleiderer, W.
Synthesis 1987, 6, 535.
11. Kvasyuk, E. I.; Kulak, T. I.; Zaitseva, G. V.; Mikhailo-
pulo, I. A.; Charubala, R.; Pfleiderer, W. Tetrahedron
Lett. 1984, 25, 3683.
12. Himmelsbach, F.; Schulz, B. S.; Trichtinger, T.; Charu-
bala, R.; Pfleiderer, W. Tetrahedron 1984, 40, 59.
13. Ti, G. S.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc.
1982, 104, 1316.
14. Eppstein, D. A.; Marsh, Y. V.; Schryver, B. B.; Larsen, M.
A.; Barnett, J. W.; Verheyden, J. P. H.; Prisbe, E. J. J.
Biol. Chem. 1982, 257, 13390.
15. Van Wijk, J.; Haasnoot, C. A. G.; de Leeuw, F. A. A. M.;
Huckriede, B. D.; Westra Hoekzema, A. J. A.; Altona, C.
PSEUROT 6.3; Leiden Institute of Chemistry, Leiden
University: Leiden, The Netherlands, 1999.
16. Mikhailopulo, I. A.; Pricota, T. I.; Sivets, G. G.; Altona,
C. J. Org. Chem. 2003, 68, 5897.