Binding of actinomycin C1 (D) and actinomin to base-modified oligonucleotide duplexes with parallel chain orientation

Article abstract of DOI:10.1016/j.bmc.2006.02.002

The binding of actinomycin D (C₁, 1) and its analog actinomin (2) was studied on base-modified oligonucleotide duplexes with parallel chain orientation (ps) and with anti-parallel chains (aps) for comparison. Actinomycin D binds not only to aps

Full text of DOI:10.1016/j.bmc.2006.02.002

Bioorganic & Medicinal Chemistry 14 (2006) 4089–4100
Binding of actinomycin C1 (D) and actinomin to
base-modified oligonucleotide duplexes with
parallel chain orientation
Hong Li,a Xiaohua Peng,a,b Peter Leonardb and Frank Seelaa,b,
*
aLaboratorium fu¨r Organische und Bioorganische Chemie, Institut fu¨r Chemie, Universita¨t Osnabru¨ck,
Barbarastraße 7, 49069 Osnabru¨ck, Germany
bCenter for Nanotechnology, Gievenbecker Weg 11, 48149 Mu¨nster, Germany
Received 4 November 2005; revised 1 February 2006; accepted 3 February 2006
Available online 24 February 2006
Abstract—The binding of actinomycin D (C1, 1) and its analog actinomin (2) was studied on base-modified oligonucleotide duplexes
with parallel chain orientation (ps) and with anti-parallel chains (aps) for comparison. Actinomycin D binds not only to aps duplex-
es containing guanine–cytosine base pairs but also to those incorporating modified bases such as 7-deazaguanine or its 6-deoxo
derivative. For this, novel phosphoramidites were prepared. The new building block of 7-deaza-20-deoxyguanosine is significantly
more stable than the one currently used and allows normal oxidation conditions during solid-phase oligonucleotide synthesis. Acti-
nomycin binds weakly to ps duplexes containing guanine–isocytosine base pairs but not to ps-DNA incorporating pairs of isogua-
nine–cytosine residues. On the contrary, the actinomycin D analog actinomin, which contains positively charged side chains instead
of the chiral peptide rings, is strongly bound to both ps- and aps-DNA. Guanines, isoguanine, as well as other 7-deaza derivatives
are accepted as nucleobases. Apparently, the pentapeptide lacton rings of actinomycin do not fit nicely into the groove of ps-DNA
thereby reducing the binding strength of the antibiotic while the groove size of ps-DNA does not affect actinomin binding notably.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
chain orientation. It functions as an inhibitor of tran-
scription both in vivo and in vitro, and results in anti-
cancer activity, in particular against Hodgkin’s
disease.9–11 The binding mode of actinomycin C1 has
been shown to be intercalative.12 The chromophore in-
serts between the DNA base pairs, while the two penta-
peptide lacton rings rest in the minor groove.13–15 In
general, the binding of 1 to DNA is sequence specific
and requires dG–dC base pairs. Single crystal X-ray
analysis of the DNA/drug complex shows that the struc-
tural basis for the dG–dC base pair selectivity is the for-
mation of hydrogen bonds in the minor groove, between
the 2-amino group or nitrogen-3 of two guanine resi-
dues, respectively, with the carbonyl oxygen atoms or
amide groups of L-threonine residues of the pentapep-
tide lacton rings.16–18 In this sense, the actinomycin–
DNA complex can serve as a model for the highly spe-
cific conformational adaptation occurring between
nucleic acids and peptides or proteins. As the resulting
actinomycin–DNA complex occupies a four base pair
duplex motif, the flanking residues to the binding site
can modulate binding strength and the kinetics of
binding.19,20 Recently, the knowledge of actinomycin
Parallel stranded (ps) DNA can be constructed from oli-
gonucleotides incorporating isoguanine–cytosine or
guanine–isocytosine base pairs instead of the canonical
guanine–cytosine pair.1–3 The dA–dT pair shows ambig-
uous base pairing and is accepted in anti-parallel strand-
ed (aps) DNA as well as in ps-DNA.4 In comparison,
the ps-DNA contains grooves of about equal size.4
Thus, the molecular recognition of ps-DNA by interca-
lating drugs or dyes might be different from DNA with
anti-parallel chain orientation. As ps-DNA can offer
new opportunities to design unique hybridization probes
or antisense constructs,5–8 the binding selectivity of
ligands (dyes, antibiotics, etc.) is of interest.
The antibiotic actinomycin C1 (D; 1) is a widely studied
ligand which binds to duplex DNA with anti-parallel
Keywords: Actinomycin D; Actinomin; Base-modified nucleotides;
Binding; Parallel DNA.
*
Corresponding author. Tel.: +49 0 541 969 2791; fax: +49 0 541 969
2370; e-mail: Frank.Seela@uni-osnabrueck.de
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.02.002

4090
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
Meval
Sar
Meval
Sar
N(C2H5)2 N(C2H5)2
O
Pro
D Val
Pro
D Val
O
(CH2)2
NH
(CH2)2
NH
H2N
NH2
Thr
NH
CO
Thr
NH
CO
CO
CO
N
C2H5
Br
N
NH2
N
NH2
O
O
O
CH3
CH3
6
4
O
CH3
CH3
1
2
3
O
O
4
NH2
NH2
R
7
N
CH3
5
3
1
HN
5
N
HN
H2N
N
HN
H2N
N
N7
N
N9
N
N
H2N
HO
N
O
N
O
N
1
3
O
O
O
O
O
HO
HO
HO
HO
HO
HO
HO
HO
HO
8a : R = H
b : R = Br
c : R = Cl
4
5
7
6
Systematic numbering
Purine numbering
Scheme 1. Structures of compounds 1–8.
binding to the observed dG–dC sequence preference of
double-stranded DNA was extended to single-stranded
DNAs.21–24 Single-stranded DNA can form fold back
structures resulting in hairpins, which bind actinomy-
cin.18,24 As the phenoxazone chromophore of actinomy-
cin intercalates nicely between the tandem G Æ T
mismatches, those hairpins containing G Æ T mismatches
also bind actinomycin.24
2. Results and discussion
2.1. Synthesis of monomers
For the synthesis of phosphoramidites 11 and 14, the
corresponding nucleosides 531 and 1232 were used as pre-
cursors. The amino group of compound 5 was protected
with dimethylaminomethylidene group using dimethyl-
formamide dimethylacetal in MeOH. Upon treatment
with water the protecting group was degraded forming
the 2-formyl derivative 9. The latter was converted into
This manuscript reports on the binding of actinomycin
to parallel DNA. The parallel DNA contains isoGd–
dC or isoCd–dG pairs along with reverse Watson Crick
dA–dT pairs. Also the replacement of dG-residues by its
7-deazapurine derivative c7Gd (4) or the 6-deoxo analog
5 and of isoGd (7) by the 7-deazapurine derivatives
8a–c is investigated using the modified bases as chemical
probes. The binding of actinomycin to ps duplexes is
compared to those with anti-parallel chain orientation.
N
N
O
N
N
H2N
HO
N
HC N
H
N
i)
ii)
O
O
HO
HO
N
HO
The studies were extended to the binding properties of
the actinomycin D analog actinomin (2). This molecule
represents a mimic of actinomycin in its ability to bind
to DNA.9,25,26 The pentapeptide lacton rings are re-
placed by aminotriethylamine residues which are posi-
tively charged under neutral conditions. Thus, the
chiral recognition of the peptide moiety is mimicked
by electrostatic forces formed with the negatively
charged DNA phosphodiester backbone. Actinomin
derivatives form dimers in aqueous solution27 as it was
observed for actinomycin D.28,29 Within our studies vis-
ible and UV-spectroscopy, CD spectra, and fluorescent
dye intercalator displacement (FID)30 of ethidium bro-
mide 3 as well as Tm-measurements are used to analyze
the chiral and the non-chiral recognition of compounds
1 and 2 to ps- and aps-DNA.
5
9
N
O
O
HC N
N
N
HC N
H
N
N
iii)
H
O
O
(MeO)2TrO
(MeO)2TrO
HO
10
O
P
NCH2CH2CO
N(iPr)2
11
Scheme 2. Synthesis of compound 11. Reagents and conditions: (i) a—
N,N-dimethylformamide dimethylacetal, MeOH, 50 ꢁC, 24 h; b—H2O,
30–40 ꢁC, 72 h; (ii) 4,40-dimethoxytriphenylmethyl chloride, anhydrous
pyridine, rt; (iii) 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite,
N,N-diisopropylethylamine, CH2Cl2.

H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
4091
the 50-O-DMT-derivative 10 under standard conditions.
Phosphitylation of 10 in anhydrous CH2Cl2 in the pres-
ence of iPr2EtN with 2-cyanoethyl-N,N-diisopropyl-
chlorophosphoramidite furnished the phosphoramidite
11 (Scheme 2).
All new monomers were characterized by elemental anal-
yses as well as by 1H and 13C NMR spectra (Table 1).
2.2. Sequence selection and duplex stability in the
presence or absence of dyes
Also a new phosphoramidite of 7-deaza-20-deoxygua-
nosine (4) was prepared which does not show draw-
backs of the corresponding phosphoramidite used
nowadays. This building block is significantly more sta-
ble against oxidation and allows normal oxidation con-
ditions in solid-phase oligonucleotide synthesis. The
sensitivity of 4 is reduced by the protection of the 6-
oxo group of compound 12 with a diphenylcarbamoyl
residue generating a fully aromatic nucleobase. For this
purpose nucleoside 12 was treated with diphenylcarba-
moyl chloride in pyridine to yield the 6-oxo-protected
nucleoside 13. Phosphitylation with 2-cyanoethyl-N,N-
diisopropylchlorophosphoramidite furnished phospho-
ramidite 14 (Scheme 3).
Earlier studies using natural DNA with heterogeneous
base sequences have shown that actinomycin D binding
strength is directly correlated to the dG–dC content of
duplex DNA. Crystal structures of the actinomycin D
complex showed that the binding is intercalative and
the phenoxazone moiety is inserted between two dG–
dC base pairs. These studies were extended to short
aps duplexes with defined sequence. It was observed
that neighbors next to the dG–dC base pairs influence
binding. Moreover, it was found that binding occurs
between other base pair motifs in which the intercalated
drug has only one dG–dC base pair as nearest neigh-
bor.33 As we want to be flexible in the replacements
of nucleobases around the bound phenoxazone moiety,
we selected the duplex 50-d(G-T)6 (15) Æ 30-d(C-A)6 (16)
as a reference for duplexes with anti-parallel chain ori-
entation and the duplexes 50-d(G-T)6 (15) Æ 50-d(6-
A)6(19) and 50-d(7-T)6 (20) Æ 50-d(C-A)6 (21) as those
with parallel chains. Oligonucleotide duplexes are
shown in Table 2.
O
O C NPh2
O
N
N
HN
i-BuHN
(MeO)2TrO
N
N
N
i-BuHN
i)
The aps-duplex 15 Æ 16 with a Tm value of 56 ꢁC contains
six dG–dC pairs (motif Ia, Fig. 1) and six dA–dT base
pairs. This duplex is designed as a target molecule for
actinomycin D or actinomin binding to aps-DNA.
Within this duplex the dG residues can be replaced by
7-deaza-20-deoxyguanosine (4, c7Gd) (motif Ib, duplex
17 Æ 16) and 2-amino-7-deazapurine 20-deoxyribofurano-
side (5) (duplex 18 Æ 18). In order to study the binding
selectivity on duplexes with parallel chain orientation,
the dG–dC pairs (motif Ia) of the duplex 15 Æ 16 were re-
placed by dG–iCd pairs (motif III, duplex 15 Æ 19) or
iGd–dC pairs (motif IV, duplex 20 Æ 21). Also base pairs
7:6 (motif II, aps duplex 20 Æ 25) and 8a–c: dC (motif V,
ps duplexes 22 Æ 21, 23 Æ 21, and 24 Æ 21) are introduced in
the oligonucleotides. These base pair motifs are shown
in Figure 1.
O
O
(MeO)2TrO
HO
HO
12
13
O
O C NPh2
N
N
i-BuHN
(MeO)2TrO
N
ii)
O
O
P
NCH2CH2CO
N(iPr)2
14
Scheme 3. Synthesis of compound 14. Reagents: (i) diphenylcarba-
moyl chloride, anhydrous pyridine, N,N-diisopropylethylamine; (ii)
2-cyanoethyl-N,N-diisopropylchlorophosphoramidite, N,N-diisopro-
pylethylamine, CH2Cl2.
The Tm values of all duplexes without ligands and in the
presence of ligands are shown in Table 2. From that it is
apparent that ps duplexes are less stable than those with
Table 1. 13C NMR Chemical shifts (d) of 7-deazapurine nucleosidesa
Compoundb,c
C(2)d
C(2)
C(4)d
C(6)
C(4a)
C(5)
C(5)
C(7)
C(6)
C(8)
C(7a)d
C(4)
C(10)
C(20)
C(30)
C(40)
C(50)
5
9
159.6
152.4
152.3
146.7
141.7
150.3
150.6
150.6
156.5
157.0
111.1
101.1
101.0
104.2
105.5
100.0
116.0
116.0
102.7
99.5
122.1
126.1
126.0
119.0
125.8
152.6
151.3
151.1
147.5
144.9
81.7
82.3
82.4
82.5
82.9
39.7
e
70.6
71.1
70.7
70.6
70.8
87.2
87.3
85.4
85.3
85.5
61.7
61.9
64.2
64.0
64.3
e
e
e
10
12f
13
a Measured in DMSO-d6.
b Systematic numbering.
c Purine numbering.
d Tentative.
e Superimposed by DMSO.
f See Ref. 32.

4092
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
Table 2. Tm-Values of duplexes in the presence and the absence of actinomycin (1) and actinomin (2)a,b
Duplex
Tm (ꢁC)
Without dye
56
With actinomycin
58
With actinomin
58
50-d(G-T-G-T-G-T-G-T-G-T-G-T)
30-d(C-A-C-A-C-A-C-A-C-A-C-A)
50-d(4-T-4-T-4-T-4-T-4-T-4-T)
30-d(C-A-C-A-C-A-C-A-C-A-C-A)
50-d(A-T-A-C-5-T-A-C-5-T-A-T)
30-d(T-A-T-5-C-A-T-5-C-A-T-A)
50-d(7-T-7-T-7-T-7-T-7-T-7-T)
30-d(6-A-6-A-6-A-6-A-6-A-6-A)
(15)
(16)
(17)
(16)
(18)
(18)
(20)
(25)
56
22
63
59
25
63
58
22
66
50-d(G-T-G-T-G-T-G-T-G-T-G-T)
50-d(6-A-6-A-6-A-6-A-6-A-6-A)
50-d(7-T-7-T-7-T-7-T-7-T-7-T)
50-d(C-A-C-A-C-A-C-A-C-A-C-A)
50-d(8a-T-8a-T-8a-T-8a-T-8a-T-8a-T)
50-d(C-A-C-A-C-A-C-A-C-A-C-A)
50-d(8b-T-8b-T-8b-T-8b-T-8b-T-8b-T)
50-d(C-A-C-A-C-A-C-A-C-A-C-A)
50-d(8c-T-8c-T-8c-T-8c-T-8c-T-8c-T)
50-d(C-A-C-A-C-A-C-A-C-A-C-A)
(15)
(19)
(20)
(21)
(22)
(21)
(23)
(21)
(24)
(21)
40
46
46
61
58
40
46
46
61
58
41
47
47
62
60
a Measurements were performed in 100 mM NaCl, 10 mM MgCl2, 10 mM sodium cacodylate, pH 7.0. Melting temperatures (Tm-values) were taken
from the first derivative of the melting curve (A260 vs temperature) measured from 20 to 80 ꢁC with 1 ꢁC minꢀ1 temperature increase.
b The concentration of oligonucleotide duplexes and of actinomycin or actinomin was 5.0 lM (1:1 ratio).
H
H
O
N
CH3
O
N
N
N
H
N
H
X
N
N
N
N H
O
N
O
H
H
N
N
N
N
H N
H
H
motif II: iGd - MeiCd(aps)
motif Ia: X = N (dG - dC) (aps)
Ib: X = C (4 - dC)
H
H
H
R
H N
O
O
N
O
N
N
N H
N H
O
N H
N H
O
N
N
N
N
N
N
N
N
N
O
N
H
H
N
N
N
H N
H N
Me
H
H
H
motif III: dG - m5isoCd (ps)
motif IV: iGd - dC (ps)
motif Va: R = H (8a - dC) (ps)
Vb: R = Br (8b -dC)
Vc: R = Cl (8c -dC)
Figure 1. Base pair motifs of ps or aps duplexes containing canonical and modified nucleosides.
aps chain orientation, except for those incorporating the
7-halogenated 7-deaza-20-deoxyisoguanosines which
show a higher stability than that of the reference duplex
15 Æ 16. The latter indicates that the halogen substituents
increase duplex stability most likely by better base stack-
ing and shows that the halogeno substituents are well
accommodated in the groove of parallel DNA.34 From
Table 2 it is obvious that the addition of the actinomycin
D increases the melting temperature of aps duplexes con-
taining dG–dC (duplex 15 Æ 16) or c7Gd–dC base pairs
(duplex 17 Æ 16) or those incorporating 2-amino-7-dea-
zapurine residues (duplex 18 Æ 18) by about 2–3 ꢁC. A
typical melting profile of the aps duplex 17 Æ 16 in the
presence and absence of actinomycin D is shown in
Figure 2, which indicates that the melting temperature
of the aps duplex incorporating c7Gd–dC base pairs
was increased by about 3 ꢁC after the addition of the
actinomycin D. The aps duplex containing iG:iC base
pairs (20 Æ 25) as well as ps duplexes do not show such
an increase. This clearly demonstrates that actinomycin
D can bind to aps DNA containing dG residues or its
7-deaza analogs or those incorporating 2-amino-7-dea-
zapurines, but not to those incorporating iGd:iCd base
pairs. It shows weak bind to ps DNA containing dG-
iCd base pairs but no binding to those incorporating
iGd-dC base pairs.
Different from actinomycin D its analog 2 is bound both
to aps and ps duplexes thereby increasing the Tm values
by 1–3 ꢁC (Table 2). In the case of self-complementary

H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
4093
1.02
1.00
0.98
0.96
0.94
0.92
0.90
0.88
0.86
0.84
17 Æ 16). For example, for the ps duplex 20 Æ 21, the
CD spectra did not change after the addition of com-
pound 1 (Fig. 3b), while the addition of compound 2
leads to strong spectral changes, especially at 260 nm.
Thus, the high affinity of compound 2 to aps and ps
duplexes as well as that of compound 1 to aps duplexes
is verified.
duplex 17•16
1+duplex 17•16
2.4. Fluorescent dye intercalator displacement
Fluorescence intercalator displacement (FID) is a tech-
nique for establishing the DNA binding affinity, se-
quence selectivity, and binding stoichiometry. The
method was introduced by Boger30 and was reviewed
recently.35 Currently, this protocol was applied for the
analyses of DNA-binding by peptide derived natural
products.36 In the work described next, a fluorescent
intercalator displacement assay (FID) was performed
on complementary oligonucleotide duplexes. They were
treated with ethidium bromide (3), resulting in a fluores-
cence increase upon binding. It was expected that the
bound intercalator 3 can be displaced when the duplex
shows stronger binding toward other ligands (1 or 2),
thereby leading to a decrease of the fluorescence of com-
pound 3. The extent of fluorescence decrease should be
directly related to the binding ability of compound 1 or
2. Typical results are shown in Figure 4. In the aps du-
plex 15 Æ 16, the fluorescence intensity of compound 3 is
decreased due to the addition of compound 1 or 2, result-
ing from the release of the bound 3. It can be deduced
that the complexes formed between aps duplexes 15 Æ 16
and compound 1 or 2 are more stable than that with ethi-
dium bromide 3. For the ps duplex 20 Æ 21 the situation is
different. The formation of the actinomin–DNA complex
was observed accompanied by a decreasing fluorescence
of 3 (Fig. 4b). As the fluorescence decrease was almost
identical in both cases (Fig. 4b), the binding constant
of actinomin (2) to the ps duplex 20 Æ 21 is similar to that
of aps-DNA (15 Æ 16). However, when compound 1 was
added to a mixture of the ps duplex 20 Æ 21 and com-
pound 3 the fluorescence changed only very little. The re-
sults observed by the FID assays are consistent with the
former Tm and CD-measurements.
20
30
40
50
60
70
80
90
Temperature (°C)
Figure 2. Normalized absorbance-temperature profiles of DNA melt-
ing of the duplex 17 Æ 16 in the absence of actinomycin (squares) and in
the presence of compound 1 (1:1 ratio) (circles). For details see
Section 4.
aps duplex 18 Æ 18 containing four nucleoside 5-residues,
the Tm-increase was not observed in the presence of acti-
nomin, a phenomenon which will be discussed below.
2.3. CD spectra
The binding of compounds 1 and 2 to duplex DNA was
also monitored by comparing the CD spectra of the
drug and the drug/duplex complex. As shown in Figure
3a, the aps-DNA (15 Æ 16) adopts the spectrum of a typ-
ical B-type structure. The maxima of positive and nega-
tive lobes are located at 278 and 240 nm. The molar
ellipticities of these two bands are increased by the addi-
tion of compound 1 or 2. Furthermore, the CD maxima
of drug–DNA complexes are shifted to longer wave-
length compared to the original spectra. Similar obser-
vations are made when compound 2 was bound to
other duplexes, while in the case of compound 1, such
phenomenon can only be observed with aps duplexes
containing dG:dC or c7Gd:dC base pairs (15 Æ 16 or
a
30
20
10
0
b
15
10
5
0
-10
-5
duplex 15•16
2+duplex 20•21
-20
-10
2+duplex 15•16
-30
-15
1+duplex 20•21
1+duplex 15•16
-40
-20
-25
duplex 20•21
-50
200
220
240
260
280
300
320
340
360
200
220
240
260
280
300
320
340
360
Wavelength (nm)
Wavelength (nm)
Figure 3. (a) CD spectra of the aps duplex 15 Æ 16 and in the presence of compounds 1 or 2 (1:1 ratio); (b) CD spectra of the ps duplex 20 Æ 21 and in
the presence of compounds 1 or 2 (1:1 ratio). For details see Section 4.

4094
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.2
a
b
3+duplexes (aps or ps)
3+duplexes (ps or aps)
1.0
1+3+duplex 20•21
0.8
2+3+ps duplexes 20•21
2+3+aps duplexes 15•16
0.6
1+3+duplex 15•16
0.4
0.2
0.0
540
560
580
600
620
640
660
680
700
540
560
580
600
620
640
660
680
700
Wavelength (nm)
Wavelength (nm)
Figure 4. Fluorescence emission spectra of ethidium bromide (3) bound to the duplexes 20 Æ 21 or 15 Æ 16 (a) in the presence and absence of
actinomycin D (1) and (b) of actinomin (2). The excitation wavelength is 527 nm. Conditions see Section 4.
2.5. Determination of binding constants of compounds 1
and 2 to duplexes with parallel and anti-parallel chain
orientation
from Table 2 (for details, see Section 4). With the titra-
tion of the aps duplex 50-d(G-T)6 (15) Æ 30-d(C-A)6 (16)
both dyes generate a bathochromic shift of the visible
spectra with a hypochromic change, which is consistent
with earlier findings on high molecular DNA (Figs. 5a
and b).9 This results from the intercalation of the phen-
Figure 5 shows the absorption spectral changes of com-
pound 1 or 2 by the addition of the duplexes selected
a
b
2 + aps duplex 15•16
1 + aps duplex 15•16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
-0.02
0.12
0.10
0.08
0.06
0.04
0.02
0.00
300
350
400
450
500
550
600
300
350
400
450
500
550
600
Wavelength (nm)
Wavelength (nm)
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0.12
0.10
0.08
0.06
0.04
0.02
0.00
c
d
1 + ps duplex 20•21
2 + ps 20•21
300
350
400
450
500
550
600
300
350
400
450
500
550
600
Wavelength (nm)
Wavelength (nm)
Figure 5. Visible absorption spectral changes after the addition of duplexes 15 Æ 16 and 20 Æ 21 to compounds 1 (a,c) and 2 (b,d). Conditions see
Section 4.

H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
4095
oxazone moiety into the DNA duplex.33 It was already
proven long ago with actinomycin D derivatives carry-
ing bulky t-butyl groups instead of the canonical methyl
substituents in the 4- and 6-positions of the chromo-
phore that intercalation in duplex DNA cannot oc-
cur.25,37 The titration curves show an isosbestic point
at 458 nm for compound 1 and at 464 nm for compound
2 (Figs. 5a and b). These significant changes are indica-
tive of intercalation. Similar changes of the absorption
spectra of compound 1 or 2 are observed when dG
was displaced by compound 4 or 5 to form the aps
duplex 17 Æ 16 or 18 Æ 18. However, the titration of the
aps duplex 20 Æ 25 containing isoGd–isoCd base pairs
leads to the change of the absorption spectra of actino-
min (2), while no change is observed for that of actino-
mycin D.
Table 3. Comparison of the apparent binding constants of compound
1 or 2 to aps and ps duplexes
a
Chain orientation Duplex
Kap (Mꢀ1
)
Bap
1
2
1
2
Anti-parallel
Parallel
15 Æ 16 2.1 · 106
17 Æ 16 1.3 · 106
18 Æ 18 0.4 · 106
20 Æ 25 n.b.
2.0 · 106 0.9
0.8 · 106 1.0
1.5
3.0
2.0
n.b.b
n.b.b
1.5 · 106 n.b.b 2.0
15 Æ 19 ꢁ0.1 · 106 0.4 · 106 n.dc 2.0
20 Æ 21 n.b.
22 Æ 21 n.b.
1.9 · 106 n.b.b 2.4
0.7 · 106 n.b.b 2.0
a Bap, number of binding sites per duplex.
b n.b., no binding.
c n.d., not determined.
binding constant (ꢀKap) and the intercept on the x-axis
results in the multiply of the apparent number of bind-
ing sites per duplex (Bap) and the binding constant.
The observed equilibrium constants and the number of
binding sites per duplex (Bap) are listed in Table 3. With
the aps duplex 15 Æ 16 compound 1 shows the highest
binding constant, K = 2.1 · 106 Mꢀ1; the number of
binding sites per duplex is 0.9 (one binding site out of
14 base pairs). These data are in agreement with previ-
ously reported values for compound 1 with poly d(T-
G) Æ poly d(C-A) (K = 2 1 · 106 Mꢀ1 with one binding
site out of 18 8 base pairs).33 The substitution of the
dG residues by the 7-deazapurine nucleoside 4 (duplex
17 Æ 16) or 5 (duplex 18 Æ 18) reduces the binding con-
stant but still keeping them in the similar range (Table
3). As already mentioned above, the ps duplex 15 Æ 19
showed a weak binding to compound 1 (see Table 3).
When ps duplexes (20 Æ 21, 22 Æ 21, and 23 Æ 21) were add-
ed to the solution of compound 1, respectively, the
absorption spectra always kept the same (Fig. 5c). Only
the ps duplex 15 Æ 19, which is composed of dG:isoCd
and dA:dT base pairs shows a bathochromic shift with
only a small hypochromic change (data not shown).
The weak binding of compound 1 to the ps duplex
15 Æ 19 was not detected by Tm and CD measurements
probably for sensitivity reasons. The titration experi-
ments performed with actinomin (2) and ps duplexes re-
sult in a totally different behavior. No matter which ps
duplex was used, the absorption spectra of the dye chan-
ged in a way typical for phenoxazone intercalation (for
spectra, see Fig. 5d).
To obtain quantitative binding parameters, the titration
data were collected at fixed wavelengths (440 nm for
compound 1 and 443 nm for compound 2) and were
plotted in a Scatchard graph as shown in Figure 6.
The linear region of the binding isotherm at small r val-
ues (r equals bound drug molecules per duplex) was fit-
ted to Eq. (2). The slope of the line gives the apparent
The situation is different for the binding properties of
actinomin (2). This molecule displays high affinities to
aps and ps duplexes containing dG:dC base pairs or
analogs thereof. The exemption found for the duplex
18 Æ 18 is probably due to its low thermal stability. The
binding constants of 2 are always in the range from
105 to 106 Mꢀ1 and the binding sites are about two
per duplex which is significantly higher than that for
actinomycin D.
1.4
1 + aps duplex 15•16
1.2
2.6. Discussion
1.0
0.8
0.6
0.4
0.2
0.0
The requirement of guanine residues for actinomycin D
(1) binding is well established for duplexes with anti-par-
allel chain orientation.9,15 This results from a hydrogen
bond existing between the 2-amino group of the guanine
moiety and the carbonyl oxo group of the L-threonine
residue; a second hydrogen bond is formed between
the nitrogen-3 of the guanine ring (as an acceptor) and
the proton of the amide function of the L-threonine moi-
ety (as a donor).13 During complex formation the phen-
oxazone moiety of 1 is inserted into the DNA duplex
between two dG–dC base pairs. But as reported earlier
one dG–dC base pair next to actinomycin is sufficient
to bind the antibiotic.33 This is consistent with our
observation that the aps duplex d(G-T)6 Æ d(C-A)6 shows
strong binding to compound 1. The formation of drug–
DNA complexes was verified by the increase (1–3 ꢁC) of
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
r
Figure 6. Typical binding isotherm of compound 1 to the aps duplex
15 Æ 16 at 20 ꢁC. The solid line results from a linear least-squares fit
using the Scatchard equation r/m = Kap (Bap ꢀ r), where r is the value
of bound dye per mole of duplex; m is the concentration of free drug;
Bap is the apparent number of binding sites per mole of duplex and Kap
is the apparent binding constant.

4096
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
Tm values of the respective duplexes and the change of
their CD spectra. Due to the high affinity of compound
1 to aps duplexes such as 15 Æ 16, a fluorescent dye inter-
calator displacement assay with pre-bound ethidium
bromide (3) was performed. Replacement of ethidium
bromide by the drug resulted in a strong fluorescence
decrease. The same protocols were applied to the
duplexes containing modified nucleosides. The binding
of compound 1 was found to be rather strong when
the dG-residues were replaced by the 7-deazapurine ana-
logs 4 or 5, while the aps duplexes containing isoGd–
isoCd base pairs do not bind actinomycin D due to the
absence of the 2-amino group.
cantly more stable against oxidation and allows
normal oxidation conditions in solid-phase oligonucleo-
tide synthesis.
(iii) Anti-parallel oligonucleotide duplexes containing
isoGd–isoCd base pairs or parallel duplexes containing
base pairs of isoGd–dC do not bind actinomycin due
to the absence of the guanine 2-amino group.
(iv) Actinomin binds to parallel as well to anti-parallel
DNA. It is not only bound to DNA with canonical base
pairs but shows strong binding to duplexes incorporating
modified dG- or iGd-residues, for example, 7-deazapu-
rines instead of purines. Bulky 7-halogen substituents
which were present in ps duplexes do not impair actino-
min binding.
The binding behavior of compound 1 to the correspond-
ing duplexes with parallel chain orientation is different.
Duplexes with isoGd–dC base pairs do not bind actino-
mycin D, while in ps duplexes containing dG-isoCd such
as 15 Æ 19 the binding is detectable but weak. This shows
that the dG-residues present in ps duplexes can form
hydrogen bonds with compound 1. The reduced affinity
of actinomycin D to the ps duplex 15 Æ 19 is likely caused
by steric restrictions of the groove size of ps-DNA which
has to accommodate the peptide lacton rings of the anti-
biotic molecule. In aps-DNA, the shape of the chiral
peptide lacton ring follows the spatial requirements of
the minor groove of B-DNA. This chiral recognition
has already been studied with an enantiomeric analog
of actinomycin D which does not bind to aps-DNA.38
In all binding processes, solvent molecules, associated
to the backbone of DNA or the peptide lacton rings,
have to be reorganized. It has been shown that the min-
or groove becomes shallow and wide when the chain ori-
entation changes from anti-parallel to parallel, the
groove of ps duplexes cannot well accommodate the
peptide rings of compound 1. The binding of actinomin
(2) to duplex DNA is not strongly related to the chain
orientation. As shown in Scheme 1, compound 2 has
the same phenoxazone chromophore as compound 1
but different side chains. The positively charged trieth-
ylamino groups of 2 can interact with the negative
charges of the phosphodiester backbone electrostatically
forming interactions which are obviously similar both in
ps- and aps-DNA.
4. Experimental
4.1. Synthesis of compounds 9–14
4.1.1. General. All solvents are of laboratory grade. Other
chemicals were bought from Sigma-Aldrich Chemie
GmbH (Deisenhofen, Germany). Compound 2 was pre-
pared according to Seela.25 The synthesis of compounds
4–8 is described earlier.31,39–43 The phosphoramidites of
compounds 6, 7, and 8a–c were synthesized as de-
scribed.34,43,44 The standard phosphoramidites are com-
mercial materials bought from Proligo (Hamburg,
Germany). Thin-layer chromatography (TLC) was per-
formed on TLC aluminum sheets covered with silica gel
60 F254 (0.2 mm, VWR International, Darmstadt, Ger-
many). Column flash chromatography (FC) was accom-
plished on silica gel 60 (VWR International, Darmstadt,
Germany) at 0.4 bar. NMR spectra were measured on
an Avance-250 or AMX-500 spectrometer (Bruker,
Rheinstetten, Germany); Chemical shifts (d) are in ppm
relative to internal Me4Si or external H3PO4 (31P). The J
values are given in Hz. Elemental analyses were per-
formed by the Mikroanalytisches Laboratorium Beller,
Go¨ttingen, Germany.
4.1.2. 7-(2-Deoxy-b-D-erythro-pentofuranosyl)-2-form-
ylamino-7H- pyrrolo[2,3-d]pyrimidine (9). Into a solution
of 7-(2-deoxy-b-D-erythro-pentofuranosyl)-7H-pyrrolo-
[2,3-d]pyrimidin-2-amine31 (5: 750 mg, 3.0 mmol) in
MeOH (40 mL), dimethylformamide dimethylacetal
(6.0 mL, 44.8 mmol) was added and the mixture was
stirred at 50 ꢁC for 24 h. After removal of the solvent,
the residue was redissolved in MeOH (40 mL) and two
drops of water were added. The solution was stirred at
30–40 ꢁC for 72 h and adsorbed on a small amount
(5.0 g) of silica gel. This material was loaded on the
top of a silica gel column (4· 15 cm), and the product
was eluted stepwise with CH2Cl2/MeOH (98:2,
300 mL) and CH2Cl2/MeOH (95:5, 600 mL). The prod-
uct-containing fractions were combined and evaporated
to give a colorless foam (550 mg, 66%); TLC (CH2Cl2/
MeOH 9:1): Rf 0.32; UV (MeOH): kmax 238 (19,800),
3. Conclusions
From the work described above the following conclu-
sion can be drawn:
(i) Parallel duplexes containing dG-isoCd base pairs
bind actinomycin but with a much lower affinity than
anti-parallel duplexes with dG–dC pairs. This indicates
that the interaction of ps-DNA is strongly controlled
by chiral recognition of the peptide lacton rings and
the grooves of DNA.
(ii) Actinomycin binds to aps-DNA when 20-deoxygua-
nosine is replaced by 7-deaza-20-deoxyguanosine 4 or
its 6-deoxo derivative 5. Novel phosphoramidites of 4
and 5 were prepared. The new building block of 4 does
not show drawbacks of the corresponding phospho-
ramidite used nowadays. This building block is signifi-
1
274 (4460), 303 (2800); H NMR (DMSO-d6): d 2.16–
2.25 (m, 1H, 20-H), 2.50–2.56 (m, 1H, 20-H), 3.47–3.82
(m, 2H, 50-H), 3.81–3.83 (m, 1H, 40-H), 4.35–4.37 (m,
1H, 30-H), 4.90 (t, J = 5.3 Hz, 1H, 50-OH), 5.30 (d,
J = 3.8 Hz, 1H, 30-OH), 6.54–6.62 (m, 2H, 5-H, 10-H),

H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
4097
7.65 (d, J = 3.7 Hz, 1H, 6-H), 8.80 (s, 1H, 4-H), 9.43 (d,
J = 9.8 Hz, 1H, NH), 10.82 (d, J = 9.8 Hz, 1H, COH);
Anal. Calcd for C12H14N4O4 (278.26): C 51.80, H
5.07, N 20.13; found: C 51.70, H 4.97, N 20.15.
co-evaporated three times with anhydrous pyridine
(15 mL each) and dissolved in anhydrous pyridine
(30 mL). Then, diphenylcarbamoyl chloride (dpc-Cl)
(2.5 g, 10.8 mmol) and N,N-diisopropylethylamine (DI-
PEA) (1.5 mL, 8.6 mmol) were added. After stirring at rt
for 2 h, the excess of dpc-Cl was destroyed with crushed
ice (ꢂ50 mL), the mixture was poured into 5% aq NaH-
CO3 (50 mL) and extracted with CH2Cl2 (3· 100 mL).
The combined organic layer was dried over anhydrous
Na2SO4, filtered, and evaporated. The residue was submit-
ted to FC (column 4 · 15 cm, elution stepwise with
CH2Cl2 ! CH2Cl2/acetone 9:1), yields a colorless foam
(3.9 g, 77%); TLC (CH2Cl2/acetone 9:1): Rf 0.5; 1H
NMR (DMSO-d6): d 1.06–1.08 (m, 6H, CH3), 2.25–2.30
(m, 1H, 20-H), 2.57–2.68 (m, 1H, 20-H), 2.73–2.81 (m,
1H, CH), 3.14–3.25 (m, 2H, 50-H), 3.69 (s, 6H, OCH3),
3.93–3.95 (m, 1H, 40-H), 4.40–4.42 (m, 1H, 30-H), 5.36
(d, J = 3.4 Hz, 1H, 30-OH), 6.56 (t, J = 6.7 Hz, 1H, 10-
H), 6.64 (d, J = 2.7 Hz, 1H, 5-H), 6.82 (m, 4H, arom.
H), 7.21–7.49 (m, 20H, arom. H, 6-H), 10.54 (s, 1H,
NH); Anal. Calcd for C49H47N5O8 (833.93): C 70.57, H
5.68, N 8.40; found: C 70.43, H 5.61, N 8.34.
4.1.3. 7-[2-Deoxy-5-O-(4,40-dimethoxytriphenylmethyl)-
b-D-erythro-pentofuranosyl]-2-formylamino-7H-pyrrolo[2,3-
d]pyrimidine (10). Compound 9 (355 mg, 1.28 mmol)
was co-evaporated three times with anhydrous pyridine
(3 mL each) and dissolved in anhydrous pyridine
(6.0 mL). Into this solution, 4,40-dimethoxytriphenylm-
ethyl chloride (517 mg, 1.53 mmol) was added and stir-
red at rt for 6 h. The reaction mixture was quenched
by the addition of MeOH (1 mL) and evaporated to dry-
ness. The residue was dissolved in CH2Cl2 (30.0 mL)
and washed twice with 5% aq NaHCO3 (15 mL) and
once with brine (15 mL). The organic layer was dried
over anhydrous Na2SO4, filtered, and evaporated to dry-
ness. The crude product was redissolved in CH2Cl2
(2.0 mL) and submitted to FC (column 4· 10 cm, elu-
tion with CH2Cl2/MeOH, 98:2), yielding a colorless
foam (550 mg, 74%); TLC (CH2Cl2/MeOH, 95:5): Rf
0.26; UV (MeOH): kmax 243 (59,000), 276 (14,225),
300 (12,770); 1H NMR (DMSO-d6): d 2.26–2.31 (m,
1H, 20-H), 2.58–2.65 (m, 1H, 20-H), 3.15–3.21 (m, 2H,
50-H), 3.72 (s, 6H, 2MeO), 3.92–3.94 (m, 1H, 40-H),
4.38–4.41 (m, 1H, 30-H), 5.35 (d, J = 3.9 Hz, 1H, 30-
OH), 6.54–6.59 (m, 2H, 10-H, 5-H), 6.78–6.83 (m, 4H,
arom. H), 7.19–7.36 (m, 9H, arom. H), 7.47 (d,
J = 2.9 Hz, 1H, 6-H), 8.81 (s, 1H, 4-H), 9.41 (d,
J = 10.3 Hz, 1H, NH), 10.81 (d, J = 10.3 Hz, 1H,
COH); Anal. Calcd for C33H32N4O6 (580.6): C 68.26,
H 5.56, N 9.65; found: C 68.18, H 5.60, N, 9.56.
4.1.6. 7-[2-Deoxy-5-O-(4,40-dimethoxytriphenylmethyl)-b-
D-erythro-pentofuranosyl]-3,7-dihydro-2-[(2-methylpropa-
noyl)amino]-4H-pyrrolo[2,3-d]pyrimidin-4-yl diphenylcar-
bamate 30-(2-cyanoethyl N,N-diisopropylphosphorami-
dite) (14). To a solution of 13 (1.0 g, 1.2 mmol) in
anhydrous CH2Cl2 (15 mL), N,N-diisopropylethylamine
(0.6 mL,
3.4 mmol)
and
2-cyanoethyl
diiso-
propylphosphoramidochloridite (0.6 mL, 2.7 mmol)
was added. The mixture was stirred at rt for 15 min,
diluted with CH2Cl2 (50 mL) and washed twice with
5% aq NaHCO3 (50 mL), followed by brine (30 mL).
The organic layer was separated and the aqueous layer
was extracted twice with CH2Cl2 (20 mL each). The
combined organic layers were dried over anhydrous
Na2SO4, filtered and evaporated. The oily residue was
applied to FC (column 4 · 10 cm, CH2Cl2 ! CH2Cl2/
acetone 95:5), yielding a colorless foam (930 mg,
75%); TLC (CH2Cl2/acetone 95:5): Rf 0.7; 1H NMR
4.1.4. 7-[2-Deoxy-5-O-(4,40-dimethoxytriphenylmethyl)-b-
D-erythro-pentofuranosyl]-2-formylamino-7H-pyrrolo[2,3-
d]pyrimidine
30-(2-cyanoethyl)-N,N-diisopropylphosph-
oramidite (11). Into a solution of compound 10 (200 mg,
0.34 mmol) in anhydrous CH2Cl2 (3.0 mL), N,N-diiso-
propylethylamine (DIPEA) (0.1 mL, 0.57 mmol) and 2-
cyanoethyl-N,N-diisopropylchlorophosphoramidite (0.1 mL,
0.45 mmol) were added under an Ar atmosphere. The
reaction mixture was stirred at rt for 30 min, diluted with
CH2Cl2 (20 mL), and washed twice with 5% aq NaHCO3
(10 mL), followed by brine (10 mL). The organic layer
was dried over Na2SO4, filtered, evaporated, and the res-
idue was submitted to FC (column 3 · 9 cm, CH2Cl2/ace-
tone, 95:5), yielding a colorless foam (207 mg, 77%); TLC
(CH2Cl2/acetone, 9:1): Rf 0.20, 0.26; 1H NMR (CDCl3): d
1.11–1.20 (m, 12H, 4Me), 2.46 (t, J = 6.7 Hz,
OCH2CH2CN), 2.60–2.65 (m, 2H, 20-H), 3.25–3.39 (m,
2H, Me2CH), 3.58–3.72 (m, 4H, 50-H, OCH2CH2CN),
3.78 (s, 6H, 2MeO), 4.22–4.24 (m, 1H, 40-H), 4.68–4.72
(m, 1H, 30-H), 6.44 (d, J = 3.4 Hz, 1H, 5-H), 6.66 (t,
J = 6.6 Hz, 1H, 10-H), 6.77–6.82 (m, 4H, arom. H),
7.22-7.41 (m, 10H, arom. H, 6-H), 8.19 (d, J = 11.3 Hz,
1H, NH), 8.69 (s, 1H, 4-H), 9.53 (d, J = 11.4 Hz, 1H,
COH); 31P NMR (CDCl3): 150.1, 149.9.
(CDCl3):
d 0.85–1.25 (m, 18H, 6CH3), 2.43 (t,
J = 6.3 Hz, OCH2CH2CN), 2.58–2.67 (m, 2H, 20-H),
3.26–3.45 (m, 3H, Me2CH), 3.58–3.72 (m, 4H, 50-H,
OCH2CH2CN), 3.77 (s, 6H, 2MeO), 4.18–4.20 (m,
1H, 40-H), 4.68–4.70 (m, 1H, 30-H), 6.36 (d,
J = 3.4 Hz, 1H, 5-H), 6.63 (t, J = 6.5 Hz, 1H, 10-H),
6.77–6.82 (m, 4H, arom. H), 7.22–7.39 (m, 10H, arom.
H, 6-H), 7.82, 7.87 (2s, 1H, NH); 31P NMR (CDCl3):
150.0, 149.6.
4.2. Oligonucleotide synthesis and purification
The synthesis of the oligonucleotides using the protocol
of phosphoramidite chemistry was performed on an
ABI 392-08 synthesizer (Applied Biosystems, Weiters-
tadt, Germany) employing the standard conditions at
1 lmol scale following the synthesis protocol for 30-
cyanoethylphosphoramidites.45 After cleavage from
the solid support, the oligonucleotides were deprotec-
ted in 25% aqueous ammonia solution for 12–16 h at
60 ꢁC. Purification of the 50-dimethoxytrityl oligomers
was carried out on a 4 · 250 mm RP-18 (5 lm)
4.1.5. 7-[2-Deoxy-5-O-(4,40-dimethoxytriphenylmethyl)-b-
D-erythro-pentofuranosyl]-3,7-dihydro-2-[(2-methylprop-
anoyl)amino]-4H-pyrrolo[2,3-d]pyrimidin-4-yl diphenylcar-
bamate (13). Compound 1232 (3.9 g, 6.1 mmol) was

4098
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
LiChrosorb column (VWR International, Darmstadt,
Germany) with a Merck-Hitachi HPLC pump (model
655A-12) connected with a variable wavelength monitor
(model 655-A) and a controller (model D-2000). The tri-
tyl-on oligomers were chromatographed with the follow-
ing solvent gradient system [A: 0.1 M (Et3NH)OAc (pH
7.0)/MeCN 95:5; B: MeCN]: 0–3 min, 10–15% B in A, 3–
15 min, 15–50% B in A, 15–20 min, 50–10% B in A, 20–
25 min, 10% B in A, the flow rate always being 1.0 mL/
min. The solution was dried and treated with 2.5%
CHCl2COOH/CH2Cl2 for 5 min at room temperature
(rt) to remove the 4,40-dimethoxytrityl residue. The det-
ritylated oligomers were purified by reversed-phase
HPLC with the gradient: 0–25 min, 0–20% B in A with
a flow rate of 1.0 mL/min. The oligomers were desalted
with a 4 · 125 mm RP-18 (5 lm) LiChrosorb column
(VWR International, Darmstadt, Germany) and lyophi-
lized on a Speed-Vac evaporator to yield colorless solids.
The molecular masses of the oligonucleotides were deter-
mined by MALDI-TOF mass spectrometry (Table 4).
The MALDI-TOF mass spectra were recorded on a Bi-
flex-III spectrometer (Bruker, Leipzig, Germany) in the
reflector mode. The average power of the nitrogen laser
(337.1 nm) at 20 Hz was 3–4 mW (150–200 lJ/pulse)
with a delay time of 600 ns. 3-HPA (3-hydroxypicolinic
acid) was used as the matrix. Oligonucleotide concentra-
tions were determined by measuring the absorbance at
260 nm after melting (at 95 ꢁC). The extinction coeffi-
cients of the unmodified oligonucleotides were deter-
mined according to nearest-neighbor approximation
using mono- and di-nucleotide tabulated values.46 The
extinction coefficients at 260 nm of the modified
oligonucleotides were calculated from the sum of
the extinction coefficients of the monomeric 20-deoxy-
ribonucleosides corrected by the hypochromicity
(Table 4). The hypochromicity (h = [(emonomer ꢀ epolymer) ·
4.3. Tm-measurements
The UV melting of oligonucleotides was performed on a
Cary-1/1E UV/VIS spectrophotometer (Varian, Austra-
lia) equipped with a Cary thermoelectrical controller.
The experiments were carried out by increasing the tem-
perature at a rate of 1 ꢁC/min from 20 to 90 ꢁC and the
temperature was recorded every minute. The duplexes
were dissolved in buffer solution (100 mM NaCl,
10 mM MgCl2, and 10 mM sodium cacodylate, pH
7.0) using 5.0 lM of single-strand concentration. The
Tm values were obtained from the melting curves. Each
melting curve was fit to a non-self-complementary two-
state model, and the thermodynamic parameters were
obtained with the Meltwin 3.0 software package.47
4.4. UV Absorption measurement and Scatchard plots
UV spectra were recorded on a U-3200 spectrophotom-
.
eter (Hitachi, Japan), kmax in nm, e in dm3 molꢀ1 cmꢀ1
Concentrations of oligonucleotide solutions have been
determined using the molar absorption coefficients of
the dodecamers (Table 4). Concentrations of the drug
solutions were determined by measuring the absor-
bances at 440 nm (for 1, e: 24,500 Mꢀ1 cmꢀ1) and
443 nm (for 2, e: 26,100 Mꢀ1 cmꢀ1).15,25 Stock solutions
for oligonucleotides and drugs were prepared by dissolv-
ing the samples in buffer solution (100 mM NaCl,
10 mM MgCl2, and 10 mM sodium cacodylate, pH 7.0).
The titrations were carried out at 20 ꢁC by injecting 2 lL
portions of oligonucleotide duplex solutions (duplex
concentration, 1.0 lM) to the drug solution having a
concentration of 7.5 · 10ꢀ6 M. The solutions were thor-
oughly mixed and equilibrated for 5 min after each addi-
tion. The concentration of drug–DNA complex (Ck) was
calculated according to the following Eqs. 1 and 2:
(emonomer
ꢀ1] · 100%) was determined from the absor-
)
bance before and after enzymatic digestion with snake-
venom phosphodiesterase (EC 3.1.15.1, Crotallus ada-
manteus) (for details see Ref. 34). The hypochromicity
was 20% for oligonucleotides 18, 19, and 25, 10% for
compounds 17, 20, and 22, and around 5% for 23, 24
(e260 of monomers: Ad 15,400, Cd 7600, Gd 11,700, Td
8800, 4 4300, 5 4100, 6 6300, 7 7600, 8a 7400, 8b 5800,
and 8c 6400).
Ck ¼ ðeAk CT ꢀ AÞ=ðekA ꢀ ekkÞ;
ð1Þ
where eAk and ekk are the molar extinction coefficients of
free and bound drug, respectively, measured at the
wavelength of 440 nm for 1 and 443 nm for 2, A is the
absorbance of the solution, and CT is the total drug
concentration, which was determined by adding 10 ll
of the stock solution of 1 or 2 (7.5 · 10ꢀ5 M) to a highly
Table 4. Molecular masses [M+H]+ of oligonucleotides determined by MALDI-TOF mass spectrometry and extinction coefficients of
oligonucleotides
MH+ (calcd.)
MH+ (found)
e260 (Mꢀ1 cmꢀ1
)
e260 (Mꢀ1 cmꢀ1
)
a
b
Oligonucleotide
50-d(GTG TGT GTG TGT) (15)
50-d(ACA CAC ACA CAC) (16)
50-d(4T4 T4T 4T4 T4T) (17)
50-d(ATA C5T AC5 TAT) (18)
50-d(6A6 A6A 6A6 A6A) (19)
50-d(7T7 T7T 7T7 T7T) (20)
50-d(CAC ACA CAC ACA) (21)
50-d(8aT8aT8aT8aT8aT8aT) (22)
50-d(8bT8bT8bT8bT8bT8bT)(23)
50-d(8cT8cT8cT8cT8cT8cT) (24)
50-d(A6A 6A6 A6A 6A6) (25)
3739
3553
3743
3610
3638
3739
3553
3733
4207
3940
3638
3740
3553
3743
3611
3639
3740
3553
3734
4207
3941
3637
123000
138000
78600
98400
110100
70700
96200
104200
88600
110100
87500
83200
86600
104200
120200
130200
98400
138000
97200
87600
91200
130200
a Calculated extinction coefficients.
b The extinction coefficients corrected by the hypochromicity.

H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
4099
concentrated DNA (0.1 mM) solution. ekk was deter-
Supplementary data
mined by adding increasing amounts of actinomycin
D to a highly concentrated DNA solution. At this high
DNA concentration the drug added initially is quantita-
tively bound, and the extinction coefficient of the com-
plex can be calculated directly.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2006.02.002.
References and notes
Binding constants (K) were extracted from linear least-
squares fits of the experimental binding isotherms ob-
tained for the binding of 1 or 2 to DNA.
1. Seela, F.; Gabler, B.; Kazimierczuk, Z. Collect. Czech.
Chem. Commun. 1993, 58, 170.
2. Sugiyama, H.; Ikeda, S.; Saito, I. J. Am. Chem. Soc. 1996,
118, 9994.
3. Seela, F.; Wei, C.; Melenewski, A.; Feiling, E. Nucleosides
Nucleotides 1998, 17, 2045.
A þ DNA ¼ DNA ꢀ A.
The corresponding Scatchard equation used for the
analysis of the data is Eq. (2):48
4. van de Sande, J. H.; Ramsing, N. B.; Germann, M. W.;
Elhorst, W.; Kalisch, B. W.; vonKitzing, E.; Pon, R. T.;
Clegg, R. C.; Jovin, T. M. Science 1988, 241, 551.
5. Rippe, K.; Fritsch, V.; Westhof, E.; Jovin, T. M. EMBO J.
1992, 11, 3777.
r=m ¼ KapðBap ꢀ rÞ;
where r is the ratio of bound drug per mole of duplex, m
is the concentration of free drug, Bap is the apparent
number of binding sites per mole of duplex, and Kap is
the apparent binding constant.
ð2Þ
6. Marfurt, J.; Leumann, C. Angew. Chem. Int. Ed. 1998, 37,
175.
7. Fritzsche, H.; Akhebat, A.; Taillandier, E.; Rippe, K.;
Jovin, T. M. Nucleic Acids Res. 1993, 21, 5085.
8. Li, H.; Peng, X.; Seela, F. Bioorg. Med. Chem. Lett. 2004,
14, 6031.
4.5. CD spectra
The CD spectra were measured with a Jasco J-600A
spectropolarimeter (Jasco, Japan) at room temperature
using 1.0 mL water-jacketed cylindrical cells of 1-cm
path length. The oligonucleotide samples (5.0 lM single
strand concentration) were prepared in the same buffer
as used for the Tm measurements. The spectra were
recorded between 200 nm and 350 nm in intervals of
0.5 nm.
9. Muller, W.; Crothers, D. M. J. Mol. Biol. 1968, 35, 251.
¨
10. Waring, M. J. Annu. Rev. Biochem. 1981, 50, 159.
11. Phillips, D. R.; Crothers, D. M. Biochemistry 1986, 25,
7355.
12. Graves, D. E.; Velea, L. M. Curr. Org. Chem. 2000, 4, 915.
13. Kamitori, S.; Takusagawa, F. J. Mol. Biol. 1992, 225, 445.
14. Robinson, H.; Gao, Y.-G.; Yang, X.-L.; Sanishvili, R.;
Joachimiak, A.; Wang, A. H.-J. Biochemistry 2001, 40,
5587.
15. Hou, M.-H.; Robinson, H.; Gao, Y.-G.; Wang, A. H.-J.
Nucleic Acids Res. 2002, 30, 4910.
4.6. Fluorescence dye intercalator displacement (FID)
16. Liu, C.; Chen, F.-M. Biochemistry 1996, 35, 16346.
17. Sobell, H. M.; Jain, S. C. J. Mol. Biol. 1972, 68, 21.
18. Chen, F.-M.; Sha, F.; Chin, K.-H.; Chou, S.-H. Nucleic
Acids Res. 2003, 31, 4238.
19. Chen, F.-M. Biochemistry 1988, 27, 6393.
20. Chen, F.-M. Biochemistry 1992, 31, 6223.
21. Chen, F.-M.; Sha, F.; Chin, K.-H.; Chou, S.-H. Nucleic
Acids Res. 2004, 32, 271.
In a quartz cuvette (3.0 mL) containing 1.0 mL buffer
solution (100 mM NaCl, 10 mM MgCl2, and 10 mM
sodium cacodylate, pH 7.0), ethidium bromide solution
(2.0 ll) was added (0.5 mM in buffer solution of
100 mM NaCl, 10 mM MgCl2, and 10 mM sodium cac-
odylate, pH 7.0). The fluorescence (Hitachi F-4500 spec-
trophotometer, Hitachi, Japan) was measured and
normalized to 0% relative fluorescence. The duplex solu-
tion (2.0 ll, 2.0 lM single-strand concentration,
24.0 lM in base pair final concentration) was added,
and the fluorescence was measured again and normal-
ized to 100% relative fluorescence. A solution of 1 or 2
(2.0 ll, 7.5 · 10ꢀ5 M in buffer solution of 100 mM NaCl,
10 mM MgCl2, 10 mM sodium cacodylate, pH 7.0) was
added, and the fluorescence was measured after 3 min of
incubation at 20 ꢁC. Subsequent addition of 2.0 ll ali-
quots was continued until the system reached saturation
and the fluorescence remained constant by further com-
pound additions.
22. Wadkins, R. M.; Jovin, T. M. Biochemistry 1991, 30, 9469.
23. Alexopoulos, E.; Jares-Erijman, E. A.; Jovin, T. M.;
´
Klement, R.; Machinek, R.; Sheldrick, G. M.; Uson, I.
Acta Cryst. 2005, D61, 407.
24. Chin, K.-H.; Chen, F.-M.; Chou, S.-H. Nucleic Acids Res.
2003, 31, 2622.
25. Seela, F. Z. Naturforschung 1971, 26b, 875.
26. Wawra, H.; Muller, W.; Kratky, O. Die Makromolekulare
¨
Chemie 1970, 139, 83.
27. Veselkov, D. L.; Lantushenko, A. O.; Davies, D. B.;
Veselkov, A. N. Russ. J. Bioorg. Chem. 2002, 28, 342.
28. Crothers, D. M.; Sabol, S. L.; Ratner, D. I.; Muller, W.
¨
Biochemistry 1968, 7, 1817.
29. Angerman, N. S.; Victor, T. A.; Bell, C. L.; Danyluk, S. S.
Biochemistry 1972, 11, 2402.
30. Boger, D. L.; Fink, B. E.; Brunette, S. R.; Tse, W. C.;
Hedrick, M. P. J. Am. Chem. Soc. 2001, 123, 5878.
31. Seela, F.; Steker, H. Liebigs Ann. Chem. 1984, 1719.
32. Seela, F.; Driller, H. Nucleic Acids Res. 1985, 13, 911.
33. Wells, R. D.; Larson, J. E. J. Mol. Biol. 1970, 49, 319.
34. Seela, F.; Peng, X.;Li, H. J. Am. Chem. Soc. 2005, 127, 7739.
35. Tse, W. C.; Boger, D. L. Acc. Chem. Res. 2004, 37, 61.
36. Lewis, M. A.; Long, E. C. Bioorg. Med. Chem. 2006 (in
press).
Acknowledgments
We thank Dr. Yang He and Ms. Kuiying Xu for the NMR
measurements and Prof. Dr. H.-J. Hinz for fruitful discus-
sion. We appreciate the help of Mrs. Monika Dubiel and
Mrs. Elisabeth Michalek during the preparation of the
manuscript. Financial support by Roche Diagnostics
GmbH, Penzberg, is gratefully acknowledged.

4100
H. Li et al. / Bioorg. Med. Chem. 14 (2006) 4089–4100
37. Brockmann, H.; Seela, F. Chem. Ber. 1971, 104, 2751.
38. Brockmann, H.; Schramm, W. Tetrahedron Lett. 1966, 1,
2331.
39. Seela, F.; Winkeler, H.-D.; Driller, H.; Menkhoff, S.
Nucleic Acids Res. Symp. Ser. 1984, 245.
40. Tor, Y.; Dervan, P. B. J. Am. Chem. Soc. 1993, 115, 4461.
41. Seela, F.; Menkhoff, S.; Behrendt, S. J. Chem. Soc. Perkin
Trans. 2 1986, 525.
42. Seela, F.; Peng, X. Synthesis 2004, 1203.
43. Seela, F.; Wei, C. Helv. Chim. Acta 1999, 82, 726.
44. Seela, F.; He, Y.; Wei, C. Tetrahedron 1999, 55,
9481.
45. User Manual of the DNA Synthesizer. Applied Biosystems,
Weiterstadt, Germany.
46. Fasman, G.D.; CRC Handbook of Biochemistry and
Molecular Biology, 3rd ed.; Chemical Rubber Publishing
Co.: Cleveland, OH, 1975; p 589.
47. McDowell, J. A.; Turner, D. H. Biochemistry 1996, 35,
14077.
48. Scatchard, G. Ann. N.Y. Acad. Sci. 1949, 51, 660.