The adduct formation between the thioguanine-polyamine ligands and DNA with the AP site under UVA irradiated and non-irradiated conditions

Article abstract of DOI:10.1016/j.bmc.2019.115160

The AP sites are representative of DNA damage and known as an intermediate in the base excision repair (BER) pathway which is involved in the repair of damaged nucleobases by reactive oxygen species, UVA irradiation, and DNA alkylating agents. Therefore,

Full text of DOI:10.1016/j.bmc.2019.115160

Bioorganic&MedicinalChemistry27(2019)115160
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
The adduct formation between the thioguanine-polyamine ligands and DNA
with the AP site under UVA irradiated and non-irradiated conditions
Yukiko S. Abe, Shigeki Sasaki⁎
Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
The AP sites are representative of DNA damage and known as an intermediate in the base excision repair (BER)
pathway which is involved in the repair of damaged nucleobases by reactive oxygen species, UVA irradiation,
and DNA alkylating agents. Therefore, it is expected that the inhibition or modulation of the AP site repair
pathway may be a new type of anticancer drug. In this study, we investigated the effects of the thioguanine-
polyamine ligands (SG-ligands) on the affinity and the reactivity for the AP site under UVA irradiated and non-
irradiated conditions. The SG-ligands have a photo-reactivity with the A-F-C sequence where F represents a
tetrahydrofuran AP site analogue. Interestingly, the SG-ligands promoted the β-elimination of the AP site fol-
lowed by the formation of a covalent bond with the β-eliminated fragment without UVA irradiation.
AP site repair inhibition
Binding ligand
α,β-unsaturated aldehyde
Covalent bond formation
Thioguanine
1. Introduction
have the potential to become a new class of anticancer drugs. Granzhan
and co-workers suggested that the bis-naphthalene macrocycles bind to
DNA is continuously damaged by either endogenous or exogenous
factors such as reactive oxygen species, UVA irradiation, DNA alky-
lating agents, etc.1,2 Damaged nucleobases are removed in the base
excision repair (BER) pathway to form apurinic/apyrimidinic sites (AP
sites) as the intermediates.3,4 The AP site per se represents DNA damage
and generated by spontaneous hydrolysis of the nucleobase at the rate
of about 105 sites per day in human cells. As the accumulation of the AP
sites enhances the cytotoxicity and mutagenicity, the AP site is im-
mediately repaired by the highly-coordinated BER pathway.5 In other
words, the blocking of the BER pathway is expected to enhance the
cytotoxic efficacy of the AP sites.
the AP site analogues with a high affinity and inhibit their cleavage by
APE1.10 They also indicated that the bis-naphthalene ligand promoted
β -elimination of the AP site, and subsequently, formed the adduct with
the produced α, β -unsaturated aldehyde.11 A similar approach for in-
hibition of the AP site repair was reported by McCullough and Lloyd
and they developed small molecules that catalyze the DNA strand
cleavage at the AP sites.12 Their compound promotes the β - and δ
-elimination reaction at the AP sites via the formation of a Schiff base.
Both the β -eliminated products at the AP sites and the adducts with the
ligand act as the 3′-block for the subsequent polymerization in the AP
site repair system.
Over the last 20 years, many approaches have attempted to develop
the AP sites-binding molecules as therapeutic agents for cancer.6,7
Lhomme and co-workers developed intercalator-polyamine-nucleobase
conjugates, which efficiently recognized and cleaved the AP sites.8
Recently, Brabec and co-workers also demonstrated that the metallo-
helices bind to the AP sites with a high affinity and inhibit the activity
of human AP endonuclease 1 (APE1).9 APE1 is one of the most im-
portant enzymes in the BER pathway, which recognizes the AP sites and
generates a single strand break with a 5′-deoxyribose phosphate end
and a 3′-hydroxyl end as a result of the cleavage on the 5′ side of the AP
site. The generated 3′-hydroxyl end is processed either by adding sev-
eral nucleotides or by a single nucleotide replacement. The APE1 in-
hibitors increase the cytotoxicity by the accumulation of AP sites and
We have previously reported that the nucleobase-polyamine con-
jugates bind to the AP sites and cleaved DNA at the AP site by pro-
moting β -elimination.13,14 The ligands conjugating the adenine, gua-
nine, cytosine or thymine base showed a selectivity to the opposing
base of the AP sites according to the Watson-Crick base pairing. These
ligands showed no binding for the full match ODNs without AP sites.13
The ligand binding affinity was supported by nonspecific electrostatic
interactions between the polyamine part and DNA phosphate backbone.
The polyamine part was also essential for promoting β -elimination.
Thus, the A-n3-ligand conjugating the adenine with 3,3′-diaminodi-
propyl-amine effectively and selectively promoted the DNA cleavage at
the AP site with the opposing dT. On the other hand, the G-n3-ligand
did not show a selectivity in the cleavage of the DNA at the AP sites. In
⁎ Corresponding author.
E-mail address: sasaki@phar.kyushu-u.ac.jp (S. Sasaki).
https://doi.org/10.1016/j.bmc.2019.115160
Received 25 July 2019; Received in revised form 7 October 2019; Accepted 9 October 2019
Availableonline31October2019
0968-0896/©2019ElsevierLtd.Allrightsreserved.

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
this study, we attempted to improve the property of the G-n3-ligand
regarding selectivity and reactivity toward the AP site. As the thio-
guanine base is known to form a photo-crosslink with the DNA base by
UVA irradiation,15,16 we expected that the thioguanine base would
form a covalent bond with DNA having the AP site by UVA irradiation.
The 6-S of the thioguanine base has a higher nucleophilicity than 6-O of
the guanine, leading to the expectation to form an adduct with the
unsaturated aldehyde produced via the β -elimination of the AP site.
composed of the AP site with the tetrahydrofuran derivative (F) and dA,
dG, dT dC at its opposing site Z, which is located in the flanking 5′-, and
3′-sequences (5′ GCGTAX1-F-Y1ATGCG 3′/3′ CGCATX2-Z-Y2TACGC
)
5′
(Tables 1 and 2). The SG-n3- and SG-n5-ligands did not significantly
increase the Tm values of any of the fully matched sequences (Table 3).
In the same flanking (A-F-C/T-Z-G) sequences (duplexes 1–4), the SG-
n3-ligand showed the highest ΔTm value to the cytosine base at Z
(duplex 4). Although the selectivity of the SG-n3-ligand for Z = dC is
the same as that obtained with the corresponding G-ligand, its dC-se-
lectivity and stabilizing effect were slightly lower than those of the G-
n3-ligand (ΔTm: SG-n3-ligand: +5.4 °C, G-n3-ligand: +6.6 °C).13 This
may be due to the lower thermal stability of the SG/C pair than that of
the G/C pair.18 The flanking sequences of the AP site with the F/C
combination (duplexes 4–6) affected the stabilizing effect of the SG-n3-
ligand in the order of duplex 4 (A-F-C/T-Z-G) > duplex 5 (C-F-A/G-Z-
T) > duplex 6 (A-F-A/T-Z-T). This stability trend agrees with the as-
sumption based on the nearest neighbor parameters for the Watson-
Crick duplex, in which the base pair between dC and thioguanine at the
AP site is treated as a G/C base pair.19 In contrast, the SG-n5-ligand
indicated a broader selectivity. Although the stability trend is similar to
that with the SG-n3-ligand, there was no significant difference between
the sequences. Due to the flexible pentyl linker of the SG-n5-ligand, a
variety of base pairs including the SG/C Watson-Crick type base pair,
the SG/G Hoogsteen base pair, the SG/T wobble base pair etc., might
contribute to its stability of the different sequences.
2. Results and discussion
2.1. Design and synthesis of the thioguanine-polyamine conjugates
It is known that the thiocarbonyl group of the thioguanine base
reacts with the 5,6-double bond of the thymine and cytosine base by
UVA irradiation.16 Therefore, it was expected that, when the thiogua-
nine base enters the space of the AP site, the photo-crosslinking reaction
would occur by UVA irradiation between the thiocarbonyl part and the
DNA base in a closer proximity of the ligand in the AP sites.
Two ligands with different lengths of the linker between the thio-
guanine base and the 3,3′-diaminodipropylamine were synthesized
(Scheme 1). 3,3′-Diaminodipropylamine (1) was protected with the t-
butoxycarbonyl (Boc) and the Nosyl (Ns) groups in 2 steps to give 3.13
1,3-Dibromopropane or 1,5-dibromopentane was reacted with 3 to
form the triamine bromide (4), which was conjugated with 2-amino-6-
chloropurine to produce 5. The 6-chloro group of 5 was replaced by
thiourea in ethanol to yield 6, which was subjected to the deprotection
of the Ns and Boc groups to produce the desired SG-ligand (7). Each
ligand was purified as hydrochloride salts. The ligand 7 was labelled
according to the length of the spacer: SG-n3-ligand and SG-n5-ligand
conjugated with the propyl (n = 3) and the pentyl linker (n = 5), re-
spectively.
2.3. The effect of UVA irradiation on duplexes containing AP site analogue
with SG-n3- and SG-n5-ligands
The photo-crosslinking reaction was performed by the UVA irra-
diation (4 W, 365 nm) of a solution of the SG-ligand (100 μM) and the
duplex (10 μM) in a buffer containing 10 mM HEPES-NaOH (pH 7.0),
5 mM DTT and 100 mM NaCl, and followed by an HPLC analysis
(Fig. 1). DTT was added to enhance the photoreaction.20 The strand of
the duplex is labelled according to the flanking bases, such as “the AFC
strand” for 5′-d(GCGTAA FCATGCG)-3′. In the HPLC, the peaks of the
two strands of duplexes 1 and 3 are overlapped, and those of duplexes
2, 4, 5 and 6 appeared as two peaks. The HPLC charts of duplexes 1–3
did not change after a 24-hour irradiation (Fig. 1(A)–(C)). In remark-
able contrast, duplex 4 (Z = dC) produced a new peak marked with (*)
after 24 h and accompanied by a decrease in the AFC strand (Fig. 1D),
indicating the reaction selectivity to the AP site with the opposing dC.
The photoreaction slowly proceeded and the new peak appeared after
3 h, then gradually increased (Fig. S2). Duplexes 5 or 6 having different
flanking base pairs at the F/C showed no indication of a photoreaction
by the SG-n3-ligand (Fig. 1(E) and (F)). The photoreaction for duplex 4
(Z = dC) selectively took place with the SG-n5-ligand albeit with a
lower efficiency than that of the SG-n3-ligand (Fig. 1(G)). These results
indicated that the photo-reactivity depends not only on the binding to
the AP site but also the flanking bases. In the control experiment, the G-
n3-ligand conjugating the natural guanine base did not produce photo-
products (Fig. 1(H)), indicating that the thiocarbonyl part played an
important role in the photoreaction.
2.2. Effects of SG-n3- and SG-n5-ligands on the thermal stability of the
duplex ODNs containing the AP site
I It was expected that thioguanine would form a base-pair with dC
opposite of the AP site,17 and the binding property of the SG-ligands was
evaluated by measuring the melting temperature (Tm) in the absence
and presence of the SG-ligands (Table 1). The duplexes (1–6) are
The molecular mass of the new peak marked with (*) in Fig. 1D was
measured by MALDI-TOF/MS to be 4157.23, which indicated that the
structure suffered desulfurization of the adduct between the AFC strand
and the SG-n3-ligand (Fig. S3). Although isolation and determination of
the adduct structure was unsuccessful, its structure was considered as
follows. It was speculated based on the photo-reactivity of the thio-
carbonyl group that a thietane intermediate formed between the 5,6-
double bond of dC in the AFC strand and the thiocarbonyl group of the
thioguanine base, and that the thietane intermediate further proceeded
to the ring opening and the subsequent dehydrosulfurization reaction
(Chart S1).15 Such a thietane formation is possible when the thio-
carbonyl of thioguanine approaches the 5,6-double bond of dC on the 3′
side of the AP site from the major groove (Figs. S1, A and B). This
Scheme 1. The synthesis of SG-ligands. (a) i) CF3CO2Et, MeOH, −78 °C to 0 °C
ii) (Boc)2, 0 °C to rt iii) LiOH/H2O, THF, rt, 57%; (b) NsCl, Et3N, CH2Cl2, 0 °C,
75%; (c) n = 3; 1,3-dibromopropane, K2CO3, MeCN, rt, 67%, n = 5; 1,5-di-
bromopentane, K2CO3, DMF, rt, 93%; (d) 2-amino-6-chloropurine, K2CO3,
DMF, 0 °C to rt, n = 3; 82%, n = 5; 74%; (e) thiourea, EtOH, reflux, n = 3;
73%, n = 5; 80%; (f) thiophenol, K2CO3, DMF, rt, n = 3; 94%, n = 5; 91%; (g)
0.5 M HCl/MeOH, rt, n = 3; quant., n = 5; 78%.
2

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
Table 1
Increase of the UV melting temperature (ΔTm
(
SD), °C).
Duplex (X1-F-Y1 / X2-Z-Y2) (Tm °C in the absence of the ligand)
1
2
3
4
5
6
A-F-C/T-A-G
A-F-C/ T-G-G
(34.4 ( 0.2))
+1.6 ( 0.1)
+3.8 ( 1.6)
A-F-C/T-T-G
A-F-C/T-C-G
C-F-A/ G-C-T
(32.0 ( 0.4))
+3.5 ( 0.6)
+4.1 ( 0.7)
A-F-A/T-C-T
(33.6 ( 0.2))
+2.3 ( 0.5)
+2.8 ( 0.2)
(33.0 ( 0.4))
+1.2 ( 0.9)
+3.1 ( 1.0)
(30.7 ( 0.3))
+5.4 ( 0.3)
+4.2 ( 1.1)
(29.9 ( 0.3))
+1.8 ( 0.7)
+3.3 ( 1.3)
SG-n3-ligand
SG-n5-ligand
The UV-melting curves were measured in 10 mM HEPES-NaOH buffer (pH 7.0) containing 100 mM NaCl at the scan rate of 1 °C at 260 nm in the absence and presence
of the SG-ligands. [duplex ODNs]: 2 μM, [SG-ligands]: 10 μM. F denotes tetrahydrofuran as the stable AP site analogue. The melting temperature of the duplexes in the
absence of ligand is shown in the parenthesis. The average mean and the standard deviation, ΔTm
(
SD), were calculated based on three independent experiments.
Table 2
MALDI-TOF/MS of the synthesized ODNs.
Sequence
5′-d(GCG TAX1 F Y1AT GCG)-3′
Calcd. [M-H]−
Found
A-F-C
C-F-A
A-F-A
5′-d(GCG TAA F CAT GCG)-3′
5′-d(GCG TAC F AAT GCG)-3′
5′-d(GCG TAA F AAT GCG)-3′
3847.7
3847.7
3871.7
3847.6
3847.7
3871.7
Table 3
Melting temperature of full match duplexes and ΔTm values obtained using SG-
ligands.
+
SG-n3-ligand
+
SG-n5-ligand
X1-N-Y1 / X2-N-Y2
A-T-C/T-A-G
A-C-C/T-G-G
A-A-C/T-T-G
A-G-C/T-C-G
C-G-A/G-C-T
A-G-A/T-C-T
Tm (°C)
ΔTm
(
SD) (°C)
ΔTm
(
SD) (°C)
52.0 ( 0.1)
54.8 ( 0.0)
52.1 ( 0.1)
48.5 ( 0.1)
55.2 ( 0.1)
50.3 ( 0.1)
+0.5 ( 0.2)
+0.5 ( 0.2)
+0.2 ( 0.3)
+0.7 ( 0.2)
+0.4 ( 0.2)
+0.7 ( 0.2)
+0.6 ( 0.1)
+1.0 ( 0.2)
+0.7 ( 0.3)
+0.6 ( 0.1)
+0.9 ( 0.2)
+0.6 ( 0.1)
Duplexes are 5′-d(GCG TA X1 N Y1AT GCG)-3′ and 3′-d(CGC ATX2 N Y2TA
CGC)-5′. Each melting curve was measured in the absence and presence of
10 μM ligand, at a concentration of 2 μM duplex in the buffer containing 10 mM
HEPES-NaOH (pH 7.0), 100 mM NaCl, respectively.
speculation may account for the fact that the SG-n3-ligand did not form
a photo-adduct for the CFA strand, in which the thiocarbonyl of thio-
guanine is in an opposite direction to the 5,6-double bond of dC on the
́
5 side of the AP site (Figs. S1, C and D).
Fig. 1. HPLC analysis of the reaction between the SG-n3-ligand and (A) duplex
1, (B) duplex 2, (C) duplex 3, (D) duplex 4, (E) duplex 5, (F) duplex 6. (G) The
reaction between the SG-n5-ligand and duplex 4. (H) The reaction between G-
2.4. The DNA cleavage and adduct forming reaction of the AP sites with the
SG-n3- and SG-n5-ligands
n3-ligand and duplex 4. (5′-d(GCG TAX1 F Y1AT GCG)-3′/ 3′-d(CGC ATX2
Z
Y2TA CGC)-5′, F = tetrahydrofuran.
The SG-ligands were next subjected to investigation for the DNA
cleavage and adduct forming reaction of the AP site. The single-
stranded 5′ FAM-DNA1(U) containing the uracil residue was used to
produce the AP site. The DNA1(U) was annealed with DNA2(X), which
was treated with uracil-DNA glycosylase to create the DNA1(AP)/
DNA2(X) duplex having an AP site. The cleaved fragments of 5′ FAM-
DNA1(AP) were analyzed by gel-electrophoresis, which were visualized
by FAM fluorescence. The reactions were performed using the SG-n3-
and SG-n5-ligands and all four AP/X combinations (X = dA, dG, dT, dC)
and all the gel images are shown in Fig. S4. Fig. 2 illustrates the typical
results using gels obtained with the AP/C combination and the control
experiment using the G-n3-ligand conjugating a guanine base. Fig. 2
also summarizes the time course of the reaction of all combinations.
The plausible reaction scheme of the SG-ligands with DNA via β - and δ-
elimination of the AP site is summarized in Chart 1.21–23
DNA4b.24 As a low selectivity in the DNA cleavage was observed for the
opposing base at the AP site with the G-n3-ligand in a previous study,14
the effects of the flanking bases were not investigated in this study.
Fig. 2A illustrates the time course of the reaction using the SG-n3-ligand
and DNA with the AP/C combination. The fast-moving bands observed
at 30 min correspond to the β -eliminated product DNA4 and the faster-
moving bands became larger after a longer reaction time that indicated
the δ -eliminated product DNA7. After 1 h, the slower-moving bands
increased. Considering that the slower moving bands were not observed
in the presence of the G-n3-ligand (Fig. 2B), and that these bands in-
creased as the β -elimination bands of DNA4 decreased, this slower
band was postulated to be due to the adduct of the SG-n3-ligand as
shown in DNA8. The time course of the ratio of the DNA fragments
observed in Fig. 2A is summarized in Fig. 2D, showing that DNA4 was
initially formed, followed by the formation of DNA8 at a faster rate and
DNA7 at a slower rate. Fig. 2E shows the time course of the reaction of
the SG-n3-ligand with the AP/A combination. In this case, the DNA8
formation was much slower and DNA4 remained. In contrast, for the
The AP site of DNA1 is in equilibrium with the open-chain aldehyde
DNA3, which undergoes β -elimination of its 3′-O-phosphate by the
attack of the amine base of the ligand to produce DNA4 and DNA5.
DNA4a is in equilibrium with DNA6, which produced DNA7 via δ
-elimination.14 DNA4a is also in equilibrium with the hydrated form
3

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
Fig. 3. The HPLC analysis of the reaction between the SG-ligands or G-n3-ligand
and FAM-DNA1(AP)/DNA2(C) duplex (top) and the MS spectra of each peak 3
(bottom). HPLC condition: column; Waters × Bridge C18 3.5 μm
4.6 × 150 mm, column oven; 50 °C, monitor; 254 nm, flow rate; 1 mL/min,
solvents; (A) 0.1 M TEAA buffer (pH 7.0), (B) MeCN, B conc.; 5–18%/20 min.
The product ratios obtained after a 6-hour treatment using all AP/X
combinations with the SG-n3-ligand and SG-n5-ligand are summarized
in Fig. 2H and I, respectively, Fig. 2H shows that the adduct selectivity
for the opposing base at the AP site was low. In contrast, Fig. 2I clearly
displays a higher ability of the SG-n5-ligand for the adduct formation of
DNA8. In particular, the pyrimidine bases, dC and dT, produced DNA8
in a higher ratio than the purine bases. The flexible pentyl linker of the
SG-n5-ligand may allow appropriate movements for the thioguanine to
attach the β -eliminated DNA4 at the nick space formed by the DNA1
cleavage, for which an opposing pyrimidine base may form more space
than a purine base.
Fig. 2. The reaction analysis of the DNA cleavage in the presence of the SG-
ligand or G-n3-ligand at interval times (A)–(C). The bands of the gel-images
were quantified and their ratios (%) are plotted versus time (D)–(G). The
fragment ratios observed after 6 h for all AP-X combination are summarized in
(H) and (I). The reaction was performed using 100 nM duplexes at 37 °C in the
presence of 10 μM SG-ligand and the products were resolved by 12% denaturing
polyacrylamide gel electrophoresis containing 8 M urea at 25 mA for 40 min.
The gel was visualized by FAM fluorescence using LAS-4000.
To confirm the adduct formation, we also analyzed the reaction by
HPLC using the duplex formed with DNA1(AP) and DNA2(C) (Fig. 3).
All the peaks were isolated and their structures were assigned based on
the molecular masses determined by MALDI-TOF/MS (Table 4 and Fig.
S5). After a 24-hour treatment, DNA5 (MS 5969.87) corresponding to
the fragment formed via β -elimination of DNA1 and DNA2 gave the
major peaks (Fig. 3A). Although the δ -eliminated fragment DNA7, β
-eliminated fragment DNA4 and the adduct DNA8 with the SG-n3-li-
gand were not separated by HPLC, their formations were evidenced by
the mass spectrum of Peak3 (Fig. 3A, Table 4 and Fig. S5). In the re-
action with the SG-n5-ligand, DNA4, 7 or 8 gave separated peaks of the
SG-n3-ligand and their formations were clear in the mass spectrum
(Fig. 3B). In a control experiment using the G-n3-ligand, only DNA4 and
7 were detected in the mass spectrum (Fig. 3C), indicating that the
Table 4
MALDI-TOF/MS of the isolated peaks 1–3.
Peak
DNA
Calcd. [M-H]-
Found
X = C, SG-n3-ligand
X = C, SG-n5-ligand
1
2
3
DNA5
5970.0
9027.5
3813.2
3713.6
4149.9
5970.0
9027.5
3813.2
3713.6
4177.9
5970.0
9027.5
3813.2
3713.6
5969.9
9027.4
3830.2a
3713.9
4150.4
DNA 2(C)
DNA4
Chart 1. The postulated products via β -, δ -elimination in the presence of the
SG-ligands. The β -eliminated product 4a is in equilibrium with the hydrated
4b.24 The aldehyde may be in equilibrium also with the 5-membered form and
the hydrated form.
DNA7
DNA8
1
2
3
DNA5
5970.2
DNA2(C)
DNA4
9024.6
a
3830.8
3713.6
4179.1
DNA7
reaction of the SG-n5-ligand with the AP/C combination, DNA8 formed
at a faster reaction rate (Fig. 2C), in which the adduct formation pro-
ceeded without accumulation of the β -eliminated product DNA4 as
shown in Fig. 2F. In the reaction of the SG-n5-ligand with the AP/A
combination, the DNA8 formation was slower, and DNA4 remained to
some extent (Fig. 2G).
DNA8
X = C, G-n3-ligand
a
1
2
3
DNA5
5970.1
9028.5
3831.8
3717.4
DNA2(C)
DNA4
a
DNA7
DNA4 was detected as a hydrate.24
4

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
thioguanine is the major contributor of the adduct formation. The re-
sults that the guanine base did not form the adduct in this study
(Figs. 2B and C) indicate that the 2-amino group of the guanine base did
not participate in the adduct formation.25,26
reaction mixture was stirred for 1 h at 0 °C, then poured CH2Cl2
(90.0 mL), washed with 1 M KH2PO4 aq. (100 mL), sat. aq. NaHCO3
(100 mL), and brine (100 mL). The organic extracts were dried over
anhydrous Na2SO4, filtered and evaporated to afford an orange viscous
crude product (18.3 g), which was purified by column chromatography
(siliga gel, CHCl3 : EtOAc = 5 : 1 to 3 : 1) to yield a pale yellow foam
product 3 (14.5 g, 28.0 mmol, 75%). 1H NMR (500 MHz, DMSO) δ ppm:
8.05 (1H, bs), 8.01–7.95 (2H, m), 7.88–7.84 (2H, m), 6.72 (1H, bs),
3.08 (2H, t, J = 7 Hz), 3.03 (2H, t, J = 7 Hz), 2.88–2.84 (4H, m), 1.60
(2H, bs), 1.52–1.49 (2H, m), 1.37 (9H, s), 1.34 (9H, s). 13C NMR
(125 MHz, DMSO) δ ppm: 155.5, 154.6, 147.8, 134.0, 132.6, 129.4,
124.4, 78.5, 77.5, 44.2, 40.6, 37.6, 28.6, 28.2, 28.0. HR ESI-MS m/z:
3. Conclusion
We have investigated the reactivity of SG-ligands for the duplex
having the AP site and demonstrated that the SG-ligands form the
photo-adduct in a selective manner for the opposing dC. It was also
found that the SG-ligands promote the β -elimination of the AP sites
followed by the adduct formation of the β -eliminated DNA fragment
with the SG-ligands without UVA irradiation. The SG-n5-ligand is su-
perior to the SG-n3-ligand from the viewpoints of reaction efficiency as
well as the selectivity to the opposing dC and dT. It was speculated that
the thioguanine in the nick formed by β -elimination is brought into
close proximity with the α, β -unsaturated aldehyde due to the flexible
pentyl linker. Accordingly, the SG-ligands displayed an interesting ad-
duct forming ability at the AP site, and are expected to enhance the
cytotoxicity of a DNA-acting anticancer drug through a function as the
3′-block for the repair of AP sites and subsequent polymerization.
517.231 [M + H]+ (517.233 calculated for C22H36N4O8S). IR cm−1
1681, 1542, 1420, 1366, 1251, 1166.
:
4.2.3. tert-butyl
(3-((N-(3-bromopropyl)-2-nitro-phenyl)sulfonamido)-
propyl)(3-((tert-butoxy-carbonyl)amino)propyl)carbamate (4a (n = 3))
A solution of 3 (11.0 g, 21.3 mmol) in dry MeCN (39.0 mL) was
added to a suspension of K2CO3 (14.7 g, 107 mmol) and 1, 3-di-
bromopropane (10.9 mL, 107 mmol) under an argon atmosphere at
0 °C. The reaction mixture was stirred at 50 °C overnight, then filtered.
The filtrate was evaporated and dissolved in CHCl3 (100 mL), washed
with H2O (50.0 mL) and brine (50.0 mL). The organic extracts were
dried over anhydrous Na2SO4, filtered and evaporated to afford a clear
pale yellow oil crude product (25.0 g), which was purified by column
chromatography (siliga gel, CHCl3 : EtOAc = 10 : 1 to 6 : 1) to yield a
pale yellow viscous product 4a (n = 3) (9.10 g, 14.3 mmol, 67%). 1H
NMR (500 MHz, CD3Cl3) δ ppm: 8.03 (1H, d, J = 7 Hz), 7.73–7.67 (2H,
m), 7.63 (1H, d, J = 7 Hz), 3.44 (2H, t, J = 7 Hz), 3.38 (2H, t,
J = 6 Hz), 3.31 (2H, t, J = 8 Hz), 3.21 (2H, t, J = 6 Hz), 3.15 (2H, bs),
3.07 (2H, bs), 2.11 (2H, quint., J = 6 Hz), 1.80 (2H, quint., J = 7 Hz),
1.63 (2H, bs), 1.45 (9H, s), 1.43 (9H, s). 13C NMR (125 MHz, CDCl3) δ
ppm: 156.2, 148.2, 133.9, 133.2, 131.9, 131.2, 124.4, 79.1, 45.9, 44.5,
37.5, 31.4, 30.0, 28.6, 27.4. HR ESI-MS m/z: 637.190, 639.188
4. Experimental
4.1. Method
Melting points were determined using a Stuart Scientific MELTING
POINT APPARATUS SMP 3. 1H spectra were recorded at 500 MHz on a
Bruker Ascend 500 using CDCl3 or DMSO as a solvent. 13C NMR spectra
were recorded at 125 MHz on a Bruker Ascend 500 using CDCl3 or
DMSO as a solvent. Chemical shifts are reported in ppm, in δ units. ESI-
MS spectra were recorded on a Bruker micrOTOF II -SKF instrument. IR
spectra were recorded on a PerkinElmer Spectrum One FT-IR spectro-
meter.
[M + H]+ (637.190, 639.188 calculated for C25H41BrN4O8S). IR cm−1
2975, 1686, 1545, 1418, 1366, 1160.
:
4.2. Synthesis of the ligands
4.2.1. tert-butyl (3-aminopropyl)(3-((tert-butoxycarbonyl)amino)-propyl)
carbamate (2)
4.2.4. tert-butyl
(3-((N-(5-bromopentyl)-2-nitro-phenyl)sulfonamido)-
propyl)(3-((tert-butoxy-carbonyl)amino)propyl)carbamate (4b (n = 5))
A solution of ethyl trifluoroacetate (8.40 mL, 70.9 mmol) in MeOH
(50.0 mL) was dropwise added over a period of 3 h a solution of 3,3′-
diaminodipropylamine (10.0 mL, 70.9 mmol) in MeOH (120 mL) at
−78 °C. The reaction mixture was stirred for 30 min at −78 °C, then
K2CO3 (2.17 g, 15.7 mmol) and a solution of 3 (2.70 g, 5.23 mmol)
in dry DMF (6.00 mL) were added to 1, 5-dibromo-pentane (3.60 mL,
26.1 mmol) under an argon atmosphere at 0 °C. The reaction mixture
was stirred at rt overnight, then poured ether (40.0 mL × 2), washed
with H2O (60.0 mL) and brine (50.0 mL). The organic extracts were
dried over anhydrous Na2SO4, filtered and evaporated to afford a clear
yellow oil crude product (8.19 g), which was purified by column
chromatography (siliga gel, CHCl3 : EtOAc = 10 : 1) to yield a yellow
viscous product 4b (n = 5) (3.20 g, 4.85 mmol, 93%). 1H NMR
(500 MHz, CD3Cl3) δ ppm: 8.00 (1H, bs), 7.71–7.66 (2H, m), 7.62 (1H,
d, J = 8 Hz), 3.35 (2H, t, J = 7 Hz), 3.31–3.27 (4H, m), 3.20 (2H, t,
J = 6 Hz), 3.13 (2H, bs), 3.07 (2H, bs), 1.85–1.77 (4H, m), 1.62–1.52
(4H, m), 1.45 (9H, s), 1.43 (9H, s), 1.41–1.39 (2H, m). 13C NMR
(125 MHz, CDCl3) δ ppm: 156.1, 148.2, 133.7, 133.6, 131.8, 130.9,
124.3, 80.1, 79.2, 47.2, 45.2, 44.5, 44.1, 37.6, 33.5, 32.2, 28.6, 27.4,
25.2. HR ESI-MS m/z: 687.207, 689.205 [M + H]+ (687.203, 689.202
calculated for C27H45BrN4O8S). IR cm−1: 2932, 1683, 1545, 1366,
1160.
raised to 0 °C.
A solution of di-tert-butyl dicarbonate (57.0 mL,
248 mmol) in MeOH (40.0 mL) was dropwise added the reaction mix-
ture, and this mixture was stirred for 4 h at rt. MeOH was evaporated
and the residue was dissolved in THF (60.0 mL), then aq. LiOH
(120 mL) was added to the solution. The reaction mixture was stirred
overnight at rt, then neutralized. CHCl3 (200 mL × 2) was poured into
the solution and washed with H2O (120 mL), brine (200 mL). The or-
ganic extracts were dried over anhydrous Na2SO4, filtered and evapo-
rated to afford a yellow viscous crude product (33.6 g), which was
purified by column chromatography (silica gel, CHCl3 : MeOH : 28%
aq·NH3 = 180 : 20 : 1 to 90 : 10 : 1) to yield a pale yellow viscous
product 2 (13.4 g, 40.6 mmol, 57%). 1H NMR (500 MHz, CDCl3) δ ppm:
5.27 (1H, bs), 3.28–3.17 (4H, m), 3.12–3.07 (2H, m), 2.68 (2H, t,
J = 7 Hz), 1.68–1.61 (4H, m), 1.45 (9H, s), 1.42 (9H, s). 13C NMR
(125 MHz, CDCl3) δ ppm: 156.2, 79.8, 79.1, 44.5, 43.8, 39.5, 37.6,
32.7, 31.8, 28.6. HR ESI-MS m/z: 332.259 [M + H]+ (332.254 calcu-
lated for C16H33N3O4). IR cm−1: 1676, 1481, 1421, 1366, 1251, 1167.
4.2.5. tert-butyl
(3-((N-(3-(2-amino-6-chloro-9H-purin-9-yl)propyl)-2-
nitrophenyl)sulfonamide)-propyl)(3-((tert-butoxycarbonyl)amino)propyl)-
carbamate (5a (n = 3))
4.2.2. tert-butyl (3-((tert-butoxycarbonyl)amino)-propyl)(3-((2-nitrophenyl)
sulfonamido)propyl)-carbamate (3)
2-Amino-6-chloropurine (2.20 g, 12.8 mmol) and K2CO3 (3.54 g,
25.6 mmol) were added to a solution of 4a (n = 3) (9.00 g, 14.1 mmol)
in anhydrous DMF (15.0 mL) under an argon atmosphere at 0 °C. The
reaction mixture was stirred at rt overnight, then poured H2O (100 mL),
extracted with CHCl3 (100 mL) and washed with brine (100 mL). The
Dry Et3N (26.0 mL, 186 mmol) and 2-nitrobenzenesulfonyl chloride
(9.10 g, 41.0 mmol) were added to a solution of 2 (12.4 g, 37.3 mmol)
in anhydrous CH2Cl2 (65.3 mL) under an argon atmosphere at 0 °C. The
5

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
organic extracts were dried over anhydrous Na2SO4, filtered and eva-
porated to afford a brown viscous crude product (15.7 g), which was
purified by column chromatography (siliga gel, CHCl3 : MeOH = 60 : 1
to 20 : 1) to yield a pale yellow foam product 5a (n = 3) (7.60 g,
10.5 mmol, 82%). 1H NMR (500 MHz, DMSO) δ ppm: 8.11 (1H, s), 7.96
(1H, d, J = 8 Hz), 7.91 (1H, d, J = 8 Hz), 7.87 (1H, t, J = 8 Hz), 7.79
(2H, t, J = 8 Hz), 6.88 (2H, s), 6.72 (1H, bs), 4.04 (2H, t, J = 7 Hz),
3.33 (2H, t, J = 8 Hz), 3.03 (2H, t, J = 7 Hz), 2.85 (2H, q., J = 7 Hz),
2.05 (2H, quint., J = 7 Hz), 1.50–1.49 (2H, m), 1.36 (9H, s), 1.33 (9H,
bs). 13C NMR (125 MHz, DMSO) δ ppm: 160. 2, 156.0, 154.5, 149.8,
148.0, 143.5, 135.1, 132.9, 131.9, 130.1, 124.9, 123.9, 79.0, 77.9,
45.7, 45.3, 44.5, 41.0, 38.1, 28.7, 28.4, 28.2. HR ESI-MS m/z: 726.283
[M + H]+ (726.280 calculated for C30H44ClN9O8S). IR cm−1: 1678,
1613, 1545, 1367, 1161.
7.85 (1H, s), 6.75 (2H, bs), 6.73 (1H, bs), 3.91 (2H, t, J = 7 Hz),
3.25–3.16 (4H, m), 3.06–3.03 (4H, m), 2.86 (2H, q, J = 6 Hz), 1.70
(2H, quint., J = 7 Hz), 1.63 (2H, quint., J = 7 Hz), 1.53–1.47 (4H, m),
1.34 (9H, s), 1.33 (9H, s), 1.16 (2H, quint., J = 8 Hz). 13C NMR
(125 MHz, DMSO) δ ppm: 174.8, 155.5, 154.6, 152.9, 147.8, 147.5,
140.5, 134.4, 132.4, 131.8, 129.6, 128.3, 124.3, 78.5, 77.5, 47.1, 45.0,
44.1, 42.6, 37.6, 28.7, 28.2, 28.0, 27.4, 23.0. HR ESI-MS m/z: 752.320
[M + H]+ (752.322 calculated for C32H49N9O8S2). IR cm−1: 1682,
1635, 1580, 1543, 1367, 1161.
4.2.9. 2-amino-9-(3-((3-((3-aminopropyl)amino)-propyl)amino)propyl)-
1,9-dihydro-6H-purine-6-thione (7a (n = 3)); SG-n3-ligand
K2CO3 (1.03 g, 7.47 mmol) and thiophenol (281 μl, 2.74 mmol)
were added to a solution of 6a (n = 3) (1.80 g, 2.49 mmol) in anhy-
drous DMF (4.20 mL) under an argon atmosphere at 0 °C. The reaction
mixture was stirred for 4 h at rt, then evaporated to afford a dark orange
viscous crude product (3.70 g), which was purified by column chro-
matography (siliga gel, CHCl3 : MeOH : 28% aq·NH3 = 90 : 10 : 1 to 40 :
8 : 1) to yield a yellow viscous product (1.26 g, 2.35 mmol, 94%). 1H
NMR (500 MHz, DMSO) δ ppm: 7.85 (1H, s), 6.75 (3H, bs), 3.99 (2H, t,
J = 7.0 Hz), 3.13–3.07 (4H, m), 2.88 (2H, q, J = 6 Hz), 1.83 (2H,
quint., J = 7 Hz), 1.57–1.54 (4H, m), 1.37 (9H, s), 1.36 (9H, s). 13C
NMR (125 MHz, DMSO) δ ppm: 174.9, 155.6, 154.7, 152.9, 147.8,
140.6, 128.3, 78.3, 77.5, 46.7, 46.0, 44.4, 44.2, 40.9, 37.6, 29.4, 28.8,
28.3, 28.1. HR ESI-MS m/z: 539.309 [M + H]+ (539.312 calculated for
C24H42N8O4S). IR cm−1: 1678, 1578, 1421, 1391, 1366, 1251, 1171.
This Nosyl deprotected 6a (n = 3) (1.25 g, 2.32 mmol) was dis-
solved in 0.5 M HCl/MeOH (46.0 mL, 23.2 mmol) and the reaction
mixture was stirred for 6 days at rt. The precipitates were collected on
the filter, and washed with cold HCl/MeOH to afford a crude product,
which was purified by precipitation from water/EtOH (2.20 mL/
40.0 mL) at −78 °C. The purified 7a (n = 3) was collected by cen-
trifuging (1,500 rpm, 10 min), and the supernatant removed using a
pippete. The precipitate was dried under vacuum to give 7a (n = 3) as
a pale yellow solid (1.00 g, quant.). m.p. 279–283 °C. 1H NMR
(500 MHz, DMSO) δ ppm: 12.70 (1H, s,), 9.44–9.40 (4H, m), 9.00 (1H,
s), 8.22 (3H, bs), 7.47 (2H, bs), 4.23 (2H, t, J = 6 Hz), 3.01–2.97 (6H,
m), 2.92–2.91 (4H, m), 2.22 (2H, quint., J = 7 Hz), 2.09 (2H, quint.,
J = 7 Hz), 2.01 (2H, quint., J = 7 Hz). 13C NMR (125 MHz, DMSO) δ
ppm: 173.4, 154.4, 146.5, 140.2, 122.2, 44.0, 43.9, 43.8, 41.5, 36.1,
25.2, 23.6, 22.2. HR ESI-MS m/z: 339.212 [M + H]+ (339.207 calcu-
lated for C14H26N8S). IR cm−1: 2972, 2767, 1645, 1618, 1594.
4.2.6. tert-butyl
(3-((N-(5-(2-amino-6-chloro-9H-purin-9-yl)pentyl)-2-
nitrophenyl)sulfonamido)-propyl)(3-((tert-butoxycarbonyl)amino)propyl)-
carbamate (5b (n = 5))
Compound 5b (n = 5) was prepared using compound 4b (n = 5)
(3.20 g, 4.81 mmol) and the other reagents described above. An orange
viscous crude product (5.66 g), which was purified by column chro-
matography (siliga gel, CHCl3 : MeOH = 60 : 1 to 20 : 1) to yield a pale
yellow viscous product 5b (n = 5) (2.50 g, 3.25 mmol, 74%). 1H NMR
(500 MHz, DMSO) δ ppm: 8.10 (1H, s), 7.98–7.94 (2H, m), 7.87 (1H, dt,
J = 2, 8 Hz), 7.82 (2H, t, J = 2, 8 Hz), 6.87 (2H, s), 6.72 (1H, bs), 4.00
(2H, t, J = 7 Hz), 3.24 (2H, t, J = 8 Hz), 3.19 (2H, t, J = 8 Hz),
3.05–3.02 (4H, m), 2.86 (2H, q, J = 6 Hz), 1.75 (2H, quint., J = 7 Hz),
1.64–1.62 (2H, m), 1.52–1.50 (4H, m), 1.36 (9H, s), 1.33 (9H, s),
1.23–1.14 (2H, m). 13C NMR (125 MHz, DMSO) δ ppm: 159.7, 155.5,
154.5, 154.0, 149.3, 147.5, 143.1, 134.4, 132.4, 131.8, 129.6, 124.3,
123.3, 78.5, 77.4, 47.1, 45.0, 44.1, 42.8, 37.6, 28.5, 28.2, 28.0, 27.4,
23.0. HR ESI-MS m/z: 754.312 [M + H]+ (754.311 calculated for
C32H48ClN9O8S). IR cm−1: 1673, 1615, 1545, 1367, 1161.
4.2.7. tert-butyl (3-((N-(3-(2-amino-6-thioxo-1,6-dihydro-9H-purin-9-yl)
propyl)-2-nitrophenyl)-sulfonamido)propyl)(3-((tert-butoxycarbonyl)-
amino)propyl)carbamate (6a (n = 3))
Thiourea (786 mg, 10.3 mmol) was added to a solution of 5a
(n = 3) (2.50 g, 3.44 mmol) in ethanol (46.0 mL) at rt. The reaction
mixture was refluxed for 2 h, then the solvent was evaporated. The
residue was dissolved in EtOAc (100 mL × 3), and washed with H2O
(150 mL). The organic extracts were dried over anhydrous Na2SO4,
filtered and evaporated to afford a yellow foam crude product (2.30 g),
which was purified by column chromatography (siliga gel, CHCl3
:
4.2.10. 2-amino-9-(5-((3-((3-aminopropyl)amino)-propyl)amino)pentyl)-
1,9-dihydro-6H-purine-6-thione (7b (n = 5)); SG-n5-ligand
MeOH = 40 : 1 to 10 : 1) to yield a yellow solid product 6a (n = 3)
(1.80 g, 2.50 mmol, 73%). 1H NMR (500 MHz, DMSO) δ ppm: 11.86
(1H, s), 7.96 (1H, dd, J = 1, 8 Hz), 7.89–7.85 (2H, m), 7.86 (1H, s),
7.79 (1H, dt, J = 1, 8 Hz), 6.75 (3H, bs), 3.95 (2H, t, J = 7 Hz), 3.31
(2H, t, J = 8 Hz), 3.23 (2H, t, J = 8 Hz), 3.05–3.02 (4H, m), 2.86 (2H,
q, J = 6 Hz), 1.99 (2H, quint., J = 7 Hz), 1.63 (2H, quint. J = 7 Hz),
1.50–1.49 (2H, m), 1.36 (9H, s), 1.35 (9H, bs). 13C NMR (125 MHz,
DMSO) δ ppm: 174.9, 155.5, 154.6, 152.9, 147.8, 147.5, 140.3, 134.6,
132.4, 131.5, 129.6, 128.3, 124.4, 78.5, 77.5, 45.2, 44.8, 44.0, 40.3,
37.6, 28.8, 28.3, 28.0, 27.1. HR ESI-MS m/z: 724.288 [M + H]+
(724.291 calculated for C30H45N9O8S2). IR cm−1: 1676, 1632, 1579,
1542, 1367, 1161.
K2CO3 (1.04 g, 7.55 mmol) and thiophenol (284 μl, 2.77 mmol)
were added to a solution of 6b (n = 5) (1.89 g, 2.52 mmol) in dry DMF
(5.00 mL) under an argon atmosphere at 0 °C. The reaction mixture was
stirred for 4 h at rt, then poured H2O (50.0 mL), extracted with EtOAc
(150 mL × 2) and washed with brine (80.0 mL). The organic extracts
were dried over anhydrous Na2SO4, filtered and evaporated to afford an
orange viscous crude product (3.13 g), which was purified by column
chromatography (siliga gel, CHCl3 : MeOH : 28% aq·NH3 = 90 : 10 : 1
to 40 : 8 : 1) to yield a white foam product (1.29 g, 2.28 mmol, 91%). 1H
NMR (500 MHz, DMSO) δ ppm: 7.88 (1H, s), 6.77 (2H, bs), 6.75 (1H,
bs), 3.93 (2H, t, J = 7 Hz), 3.14–3.07 (4H, m), 2.88 (2H, q, J = 6 Hz),
2.48–2.44 (4H, m), 1.72 (2H, quint., J = 7 Hz), 1.60–1.56 (4H, m), 1.41
(2H, quint., J = 7 Hz),
4.2.8. tert-butyl (3-((N-(5-(2-amino-6-thioxo-1,6-dihydro-9H-purin-9-yl)
pentyl)-2-nitrophenyl)-sulfonamido)propyl)(3-((tert-butoxycarbonyl)-
amino)propyl)carbamate (6b (n = 5))
1.37 (9H, s), 1.36 (9H, s), 1.23 (2H, quint., J = 8 Hz). 13C NMR
(125 MHz, DMSO) δ ppm: 174.9, 155.6, 154.7, 152.9, 147.8, 140.6,
128.3, 78.3, 77.5, 48.9, 46.5, 44.3, 42.7, 37.6, 29.1, 28.5, 28.2, 28.1,
27.7, 23.8. HR ESI-MS m/z: 567.348 [M + H]+ (567.344 calculated for
Compound 6b (n = 5) was prepared using compound 5b (n = 5)
(2.50 g, 3.25 mmol) and the other reagents describe above. A yellow
foam crude product (2.07 g), which was purified by column chroma-
tograpy (siliga gel, CHCl3 : MeOH = 40 : 1 to 20 : 1) to yield a yellow
foam product 6b (n = 5) (1.94 g, 2.58 mmol, 80%). 1H NMR (500 MHz,
DMSO) δ ppm: 11.85 (1H, bs), 7.98–7.94 (2H, m), 7.88–7.80 (2H, m),
C
26H46N8O4S). IR cm−1: 1682, 1576, 1391, 1366, 1251, 1167.
This Nosyl deprotected 6b (n = 5) (1.28 g, 2.26 mmol) was dis-
solved in 0.5 M HCl/MeOH (45.0 mL, 22.6 mmol) and the reaction
mixture was stirred for 6 days at rt. The precipitates were collected on
6

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
the filter, and washed with cold HCl/MeOH to afford a crude product,
which was purified by precipitation from water/EtOH (2.30 mL/
45.0 mL) at −78 °C. The purified 7b (n = 5) was collected by cen-
trifuging (1,500 rpm, 10 min), and the supernatant removed using a
pippete. The precipitate was dried under vacuum to give 7b (n = 5) as
a pale yellow solid (899 mg, 1.75 mmol, 78%). m.p. 265–268 °C. 1H
NMR (500 MHz, DMSO) δ ppm: 12.48 (1H, s), 9.31 (2H, bs), 9.17 (2H,
bs), 8.69 (1H, s), 8.15 (3H, bs), 7.29 (2H, bs), 4.04 (2H, t, J = 7 Hz),
3.05–2.95 (6H, m), 2.93 (2H, q, J = 6 Hz), 2.90–2.81 (2H, m), 2.07
(2H, quint., J = 7 Hz), 2.00 (2H, quint., J = 7 Hz), 1.80 (2H, quint.,
J = 7 Hz), 1.69 (2H, quint., J = 8 Hz), 1.30 (2H, quint., J = 8 Hz). 13C
NMR (125 MHz, DMSO) δ ppm: 173.4, 154.1, 146.8, 140.4, 123.5, 46.5,
44.1, 44.0, 43.9, 43.5, 36.2, 28.2, 24.8, 23.7, 22.9, 22.2. HR ESI-MS m/
4.6. Preparation of duplex ODNs containing the AP site
ODNs were purchase from Life Technologies Japan Ltd. (Tokyo,
JAPAN). The single-stranded ODN containing the uracil residue at po-
sition 11 of the 30 mer ODN (DNA1(U); 5′ FAM-d(GTT GGA GCT GUT
GGC GTA GGC AAG AGT GCC)-3′) was labeled at the 5′-end with 6-
carboxyfluorescein (6-FAM). The DNA1(U) was annealed at 70 °C for
5 min with DNA2(X) (5′ -d(GGC ACT CTT GCC TAC GCC AXC AGC TCC
AAC)-3′, X = A, G, C, T). The annealed duplex ODN was treated with
0.2 unit/μl uracil-DNA glycosylase (NEW ENGLAND BioLabs) in 50 mM
HEPES-NaOH buffer (pH 7.4), 1 mM EDTA, and 1 mM DTT at 37 °C for
15 min to create the DNA1(AP)/DNA2(X) duplex having an AP site,
which was purified by the Micro Bio-Spin® P-30 column.
z: 367.238 [M + H]+ (367.239 calculated for C16H30N8S). IR cm−1
:
2975, 1645, 1588, 1453, 1198, 1172.
4.7. Evaluation of cleavage and adduct forming ability of ODNs containing
the AP site with SG-ligands by gel-electrophoresis
4.3. Synthesis of ODNs containing the AP site analogue.
The DNA1(U)/DNA2(X) duplex (100 nM) was prepared as described
above. The sample was annealed again in 10 mM HEPES-NaOH buffer
containing 100 mM NaCl at 70 °C and pH 7.0 for 5 min, then incubated
with the SG-ligands (10 μM) at 37 °C for 5, 10, 15, 30, 60, 120, 180, 240,
or 300 min. The reaction mixture was quenched by the addition of the
loading buffer (5 M urea, 10 mM EDTA, TBE buffer, and 0.025 w/v %
xylene cyanol in final concentration) on ice. The cleaved fragments of
DNA1(AP) were resolved by 12% denaturing polyacrylamide gel elec-
trophoresis containing 8 M urea at 25 mA for 40 min. The gel was vi-
sualized by FAM fluorescence using LAS-4000. The results are sum-
marized in Fig. 2 and Figure S2. The time course of ratio of the DNA
fragments were calculated using ImageJ.
Complementary oligodeoxynucleotides (ODNs) (5′-d(CGC ATY2
Z
X2TA CGC)-3′) were purchase from Japan Bio Services Co., LTD.
(Saitama, JAPAN) or Life Technologies Japan Ltd (Tokyo, JAPAN).
ODNs containing the AP site analogue were synthesized using solid
phase phosphoroamidite chemistry on a NIHON TECHNO SERVICE H-
Series synthesizer. The cleaved oligonucleotides were purified by HPLC
on a COSMOSIL 5C18-MS-II column using a linear gradient between the
0.1 M TEAA buffer and CH3CN. The MALDI-TOF/MS was measured on a
BRUKER DALTONICS microflex-KS spectrometer with matrix con-
taining 3-hydroxy-2-picolinic acid and diammonium hydrogen citrate
solution. These data were summarized in Table 2. Oligonucleotide
concentrations were determined spectrophoto-metrically by measuring
the absorbance at 260 nm.
4.8. Evaluation of cleavage and adduct forming ability of ODNs containing
the AP site with SG-ligands and G-n3-ligand by HPLC
4.4. Evaluation of the stability of duplex ODNs containing the AP site
analogue with SG-ligands by thermal denaturation studies
The DNA1(U)/DNA2(C) duplex (22 μM) was prepared as described
above in 10 mM HEPES-NaOH buffer containing 100 mM NaCl at 70 °C
and pH 7.0 for 5 min, then incubated with the SG-ligand (220 μM) at
37 °C for 1, 3, 6, or 24 h. The reaction mixture was stored at −30 °C.
These samples were analyzed by HPLC using the following conditions.
Column: Waters X Bridge C18 3.5 μm, 4.6 × 150 mm; column oven:
50 °C; UV monitor at 254 nm; flow rate: 1 mL/min; solvents: A) 0.1 M
TEAA buffer (pH 7.0), B) MeCN, B 5% to 18%/20 min. The HPLC charts
are summarized in Fig. 3. Peaks 1–3 were isolated and their structure
were determined by MALDI-TOF/MS measurements, which are sum-
marized in Fig. S5. The DNA4 was detected as hydrated product
(+H2O).24
All duplexes melting curves were measured on a UV BECKMAN
COULTER DU800 in the absence and presence of 10 μM ligand, at a
concentration of 2 μM duplex in the buffer containing 10 mM HEPES-
NaOH (pH 7.0), 100 mM NaCl at a heat rate of 1 °C/min (from 20 °C to
80 °C) and a scan rate of 1 °C/min. As a pretreatment, all samples were
heated for 5 min at 80 °C and then gradually cooled down. Obtained
melting curves were analyzed by non-linear fitting and each melting
temperature was determined. ΔTm = Tm(+)-Tm(-), which were sum-
marized in Table 1. Table 3 summarized the ΔTm values obtained using
the full matched duplexes. The average mean and the standard devia-
tion, ΔTm
(
SD), were calculated based on three times independent
Declaration of Competing Interest
The author declare that there is no conflict of interest.
Acknowledgments
experiments.
4.5. Evaluation of the photoreaction of duplex ODNs containing the AP site
analogue with SG-ligands and G-n3-ligand by UVA irradiation
The photo-crosslinking reaction was performed by UVA irradiation
(4 W, 365 nm, Funakoshi UVGL-25) at the concentration of 10 μM du-
plex and 100 μM SG-ligand in the buffer containing 10 mM HEPES-
NaOH (pH 7.0), 5 mM DTT and 100 mM NaCl. The reaction mixtures
were irradiated at rt for 24 h, then these samples were analyzed by
HPLC using the following conditions. Column: Waters X Bridge C18
3.5 μm, 4.6 × 150 mm; column oven: 50 °C; UV monitor at 254 nm;
flow rate: 1 mL/min; solvents: A) 0.1 M TEAA buffer (pH 7.0), B) MeCN,
B 4% to 14%/15 min. (3% to 16%/15 min for Fig. 1F). The HPLC chart
were summarized in Fig. 1. The time course of photoreaction between
duplex 4 and SG-n3-ligand was shown in Fig. S1. The peak marked with
(*) was isolated by same HPLC conditions as above and measured
MALDI-TOF MS using matrix containing 3-hydroxy-2-picolinic acid and
diammonium hydrogen citrate solution.
This study was supported by a Grant-in-Aid for Scientific Research
(B) (Grant Number 15H04633 for S. S.) and a Grant-in-Aid for Young
Scientists (B) (Grant Number 26860079 for Y. A.) from the Japan
Society for Promotion of Science (JSPS).
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmc.2019.115160.
References
1. Barnett RN, Cleveland CL, Landman U. ACC Chem. Res. 2010;43:280–287.
2. Drabløs F, Feyzi E, Aas PA, et al. DNA Repair. 2004;3:1389–1407.
7

Y.S. Abe and S. Sasaki
Bioorganic&MedicinalChemistry27(2019)115160
3. Lindahl T, Karran P, Wood RD. Curr. Opin. Genet. Dev. 1997;7:158–169.
4. Friedman JI, Stivers JT. Biochemistry. 2010;49:4957–4967.
5. Baldwin MR, O’Brien PJ. Biochemistry. 2010;49:7879–7891.
6. Lhomme J, Constant J-F, Demeunynck M. Biopolymers. 1999;52:65–83.
7. Fishel ML, Kelly MR. Mol. Aspects Med. 2007;28:375–395.
8. Fkyerat A, Demeunynck M, Constant J-F, Michon P, Lhomme J. J. Am. Chem. Soc.
1993;115:9952–9959.
1998;42:109–124.
16. Brem R, Karran P. Photochem. Photobiol. 2012;88:5–13.
17. Špačková N, Cubero L, Šponer J, Orozco M. J. Am. Chem. Soc.
2004;126:14642–14650.
18. Somerville L, Krynetski EY, Krynetskaia NF, et al. J. Biol. Chem.
2003;278:1005–1011.
19. SantaLucia Jr J. Proc. Natl. Acad. Sci. USA. 1998;95:1460–1465.
20. Nishioka T, Oshiro I, Onizuka K, Taniguchi Y, Sasaki S. Chem. Pharm. Bull.
2016;64:1315–1320.
9. Malina J, Scott P, Brabec V. Nucleic Acids Res. 2015;43:5297–5306.
10. Kotera N, Poyer F, Granzhan A, Teulade-Fichou M-P. Chem. Commun.
2015;51:15948–15951.
21. Bailly V, Derydt M, Verly WG. Biochem. J. 1989;261:707–713.
22. Bailly V, Verly WG. Nucleic Acids Res. 1988;16:9489–9496.
23. Mahugh PJ, Knowland J. Nucleic Acids Res. 1995;23:1664–1670.
24. Sugiyama H, Fujiwara T, Ura A, et al. Chem. Res. Toxicol. 1994;7:673–683.
25. Dutta S, Chowdhury G, Gates KS. J. Am. Chem. Soc. 2007;129:1852–1853.
26. Johnson KM, Price NE, Wang J, et al. J. Am. Chem. Soc. 2013;135:1015–1025.
11. Caron C, Duong XNT, Guillot R, Bomberd S, Granzhan A. Chem. Eur. J.
2019;25:1949–1962.
12. Minko IG, Jacobs AC, de Leon AR, et al. Sci Rep. 2016;6:28894.
13. Abe Y, Nakagawa O, Yamaguchi R, Sasaki S. Bioorg. Med. Chem. 2012;20:3470–3479.
14. Abe SY, Sasaki S. Bioorg. Med. Chem. 2016;24:910–914.
15. Favre A, Saintomé C, Fourrey J-L, Clivio P, Laugâa PJ. Photochem. Photobiol. B Biol.
8