Strand breakage of a (6-4) photoproduct-containing DNA at neutral pH and its repair by the ERCC1-XPF protein complex

Article abstract of DOI:10.1039/c3ob00012e

The (6-4) photoproduct is one of the major UV-induced lesions in DNA. We previously showed that hydrolytic ring opening of the 5′ base and subsequent hydrolysis of the glycosidic bond of the 3′ component occurred when this photoproduct was treated with aqueous NaOH. In this study, we found that another product was obtained when the (6-4) photoproduct was heated at 90 °C for 6 h, in a 0.1 M solution of N,N′-dimethyl-1,2-ethanediamine adjusted to pH 7.4 with acetic acid. An analysis of the chemical structure of this product revealed that the 5′ base was intact, whereas the glycosidic bond at the 3′ component was hydrolyzed in the same manner. The strand break was detected for a 30-mer oligonucleotide containing the (6-4) photoproduct upon treatment with the above solution or other pH 7.4 solutions containing biogenic amines, such as spermidine and spermine. In the case of spermidine, the rate constant was calculated to be 1.4 × 10^-8 s^-1 at 37 °C. The strand break occurred even when the oligonucleotide was heated at 90 °C in 0.1 M sodium phosphate (pH 7.0), although this treatment produced several types of 5′ fragments. The Dewar valence isomer was inert to this reaction. The product obtained from the (6-4) photoproduct-containing 30-mer was used to investigate the enzymatic processing of the 3′ end bearing the damaged base and a phosphate. The ERCC1-XPF complex removed several nucleotides containing the damaged base, in the presence of replication protein A. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3ob00012e

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Strand breakage of a (6–4) photoproduct-containing
DNA at neutral pH and its repair by the ERCC1–XPF
protein complex†
Cite this: DOI: 10.1039/c3ob00012e
Norihito Arichi,‡a Junpei Yamamoto,a Chiaki Takahata,a Emi Sano,a Yuji Masuda,b
Isao Kuraokaa and Shigenori Iwai*a
The (6–4) photoproduct is one of the major UV-induced lesions in DNA. We previously showed that
hydrolytic ring opening of the 5’ base and subsequent hydrolysis of the glycosidic bond of the 3’ com-
ponent occurred when this photoproduct was treated with aqueous NaOH. In this study, we found that
another product was obtained when the (6–4) photoproduct was heated at 90 °C for 6 h, in a 0.1 M solu-
tion of N,N’-dimethyl-1,2-ethanediamine adjusted to pH 7.4 with acetic acid. An analysis of the chemical
structure of this product revealed that the 5’ base was intact, whereas the glycosidic bond at the 3’ com-
ponent was hydrolyzed in the same manner. The strand break was detected for a 30-mer oligonucleotide
containing the (6–4) photoproduct upon treatment with the above solution or other pH 7.4 solutions
containing biogenic amines, such as spermidine and spermine. In the case of spermidine, the rate con-
stant was calculated to be 1.4 × 10−8 s−1 at 37 °C. The strand break occurred even when the oligonucleo-
tide was heated at 90 °C in 0.1 M sodium phosphate (pH 7.0), although this treatment produced several
types of 5’ fragments. The Dewar valence isomer was inert to this reaction. The product obtained from
the (6–4) photoproduct-containing 30-mer was used to investigate the enzymatic processing of the 3’
end bearing the damaged base and a phosphate. The ERCC1–XPF complex removed several nucleotides
containing the damaged base, in the presence of replication protein A.
Received 4th January 2013,
Accepted 27th March 2013
DOI: 10.1039/c3ob00012e
www.rsc.org/obc
shown in Fig. 1. Although the (6–4) photoproduct is repaired
more efficiently by the nucleotide excision repair (NER)
Introduction
DNA undergoes chemical reactions with endogenous and pathway,2 it is more mutagenic than the CPD.3–5 DNA polymer-
exogenous factors. Ultraviolet (UV) light produces cross-links ase η that can incorporate two adenines opposite the CPD is
between two adjacent base moieties, and the major products not able to bypass the (6–4) photoproduct,6 whereas DNA poly-
are referred to as a cyclobutane pyrimidine dimer (CPD) and a merase ζ reportedly replicates DNA containing this photopro-
pyrimidine(6–4)pyrimidone photoproduct ((6–4) photopro- duct in an error-free manner.7
duct).1 The CPD is produced by the [2 + 2] cycloaddition
An important chemical property of the (6–4) photoproduct
between the two C5–C6 double bonds. In the case of the (6–4) is its alkali lability. Since a strand break occurs at the (6–4)
photoproduct, the four-membered ring containing an oxygen photoproduct site upon hot piperidine treatment, this property
or nitrogen atom, formed by a similar reaction, is so unstable was used to detect the formation of this type of lesion in
that it opens to produce the C6–C4-linked compound (1) DNA.8–12 We previously elucidated the mechanism of this
alkali degradation. The first reaction is the hydrolytic ring
opening of the 5′ base,13 as shown in Fig. 1. This intermediate
(2) was isolated, and the biochemical properties of DNA con-
aDivision of Chemistry, Graduate School of Engineering Science, Osaka University, 1-
taining this hydrolyzed photoproduct were analyzed.13 The
3 Machikaneyama, Toyonaka, Osaka 560-8531, Japan.
second reaction is the hydrolysis of the glycosidic bond at the
E-mail: iwai@chem.es.osaka-u.ac.jp; Fax: +81 6 6850 6240; Tel: +81 6 6850 6250
bResearch Institute of Environmental Medicine, Nagoya University, Furo-cho,
3′ component. The product without the sugar moiety (3) was
obtained even when the parent compound lacked the 3′-flank-
ing nucleotide as a leaving group.14 When the (6–4) photopro-
duct resides in DNA, the abasic site produced by the second
reaction leads to strand breakage by β,δ-elimination under
alkaline conditions.
Chikusa-ku, Nagoya 464-8601, Japan
†Electronic supplementary information (ESI) available: UV absorption spectra of
peaks i, ii, and iii, NMR spectra of compound 4, and a MALDI-TOF mass spec-
trum of peak v. See DOI: 10.1039/c3ob00012e
‡Present address: Graduate School of Pharmaceutical Sciences, Kyoto University,
46-29 Yoshida-Shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
bond of the (6–4) photoproduct was thermally more labile
than those of the normal bases under neutral conditions,
while that of the Dewar valence isomer was relatively stable.
Results
DMED treatment of a (6–4) photoproduct-containing tetramer
The tetramer d(AT(6–4)TG), which was used in our previous
study,14 was dissolved in a 0.1 M solution of DMED with the
pH value adjusted to 7.4 with acetic acid (AcOH). After
the mixture was heated at 90 °C for various periods of time,
the products were analyzed by reversed-phase HPLC. The reac-
tion rate was unexpectedly slow, but as shown in Fig. 2, two
peaks (ii and iii) were detected in addition to the starting
material (peak i) after 6 h. From the UV absorption spectrum
and the co-injection experiment, the product that yielded peak
ii was identified as 2′-deoxyguanosine 5′-phosphate, which was
released from the 3′ end of the tetramer. The other product
(peak iii) generated a UV absorption spectrum with maxima at
260 and 317 nm, while those of the parent tetramer were 256
and 327 nm, as shown in Fig. S1.† This spectrum was similar
to that observed for the product obtained by the degradation
of d(AT(6–4)TG) with NaOH, in which 2′-deoxyadenylic acid is
attached to the 5′-OH of 3,14 but the retention times of these
two compounds were different.
Fig. 1 Degradation of the (6–4) photoproduct under alkaline and neutral
conditions.
Structure of the product
The (6–4) photoproduct is converted to its Dewar valence
isomer by absorbing UV-A/B light.15–17 This isomer contains a
Dewar pyrimidinone as the 3′ base, which does not have the
326 nm absorption maximum typical of the aromatic pyrimidin-
one structure, and is reportedly more labile than the (6–4)
photoproduct under alkaline conditions.18 In our previous
study,14 the hydrolytic ring opening of the 5′ base of the (6–4)
photoproduct occurred quickly even at a low NaOH concen-
tration. However, the strand break caused by the glycosidic
bond cleavage at the 3′ component of the (6–4) photoproduct
was much slower than that observed for the Dewar isomer
when oligonucleotides containing these photoproducts were
treated with NaOH. In other words, the 5′ base of the (6–4)
photoproduct and the 3′ glycosidic bond of the Dewar isomer
are very alkali-labile.
To determine the chemical structure of this product, com-
pound 1, which was prepared by deprotection of an intermedi-
ate in the synthesis of the oligonucleotide building block of
the (6–4) photoproduct,20 was treated with the same solution
on a large scale, and NMR spectra of the product were
1
measured. The H and 31P NMR spectra revealed that one of
the sugar moieties was lost, although the phosphate remained.
The NOESY spectra indicated that the missing sugar belonged
to the 3′ component, in the same manner as the reaction
with NaOH.14 This is supported by the blue-shift of the UV
absorption maximum (from 327 nm to 317 nm) in the
A previous report showed that N,N′-dimethyl-1,2-ethane-
diamine (DMED) cleaved an abasic site-containing or UV-irra-
diated DNA at pH 7.4.19 This pH value is important because it
is a physiological pH. Although the products obtained by this
treatment and those generated with hot piperidine were com-
pared by gel electrophoresis, neither the lesion type nor the
reaction mechanism was elucidated in the case of the UV-irra-
diated DNA. In this study, we determined the chemical struc-
ture of the product and the rate constant of this reaction,
analyzed the reactions in several types of solutions, and
showed the removal of the 3′ blocking end by repair proteins,
using synthetic oligonucleotides containing the UV-induced
lesions. Our results demonstrated that the 3′-side glycosidic
Fig. 2 HPLC analysis of the solution containing d(AT(6–4)TG) in 0.1 M DMED–
AcOH (pH 7.4), after heating at 90 °C for 6 h. The chromatogram monitored at
317 nm, shown in red, is magnified by a factor of 2.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 1 13C chemical shifts
5′ Base
3′ Base
5′ Sugar
C1′
Comp. (ref.)
C2
C4
C5
C6
CH3
C2
C4
C5
C6
CH3
C2′
C3′
C4′
C5′
1 (21)
2 (13)
3 (14)
4
156.4
160.7
161.4
156.9
176.7
179.3
179.8
177.1
74.9
86.3
87.1
74.9
60.4
61.7
61.8
60.6
27.7
21.4
21.0
28.0
159.9
159.7
161.6
161.0
176.9
175.2
174.8
178.2
119.0
117.5
117.3
118.6
146.8
148.1
152.5
149.1
16.2
15.7
15.9
16.3
84.9
84.4
86.8
87.2
37.5
36.9
39.4
38.2
72.3
72.7
77.0
77.1
84.8
84.8
87.9
87.7
61.4
62.3
64.8
64.6
abovementioned experiment, which also demonstrated that
the 3′ pyrimidone ring was intact. The structure of the 5′ base
was determined by 13C NMR spectroscopy. The 13C signals
were assigned by using the HSQC and HMBC spectra, and a
difference in the chemical shifts of the 5′ base carbons was
detected between the products obtained by the reactions with
DMED–AcOH (4) and with NaOH (3), as shown in Table 1. The
values obtained for 4 in the present study were very close to
those reported for the parent compound (1).21 This ob-
servation and the mass spectrometry result indicated that the
hydrolytic ring opening of the 5′ base, which was found in the
alkali degradation of the (6–4) photoproduct,13 did not occur
under the conditions used in this study.
Treatment of a 30-mer with amine solutions
To investigate whether the reaction shown in Fig. 3 occurs in
longer oligonucleotides, a chemically-synthesized 30-mer con-
taining the (6–4) photoproduct, d(CTCGTCAGCATCT(6–4)
TCATCATACAGTCAGTG),20 was treated with 0.1 M DMED–
AcOH (pH 7.4), in the same manner as the tetramer, and the
result is shown in Fig. 4A. From the long-wavelength UV
absorption, peak v was identified with the 5′-side fragment
containing the (6–4) photoproduct at the 3′ end. This product,
d(CTCGTCAGCATCX) in which X was expected to have the
structure of 4, was purified by HPLC. Its structure including
the intact 5′ base was confirmed by MALDI-TOF mass
spectrometry, as shown in Fig. S9.† Peak vi corresponded to
the 3′-side fragment, namely, 5′-phosphorylated d(CATCATA-
CAGTCAGTG). The structure of the break site is shown in
Fig. 3. Although the conditions were the same, the strand
break apparently occurred more efficiently with the 30-mer
than with the tetramer.
Fig. 3 Strand breakage at the (6–4) photoproduct at pH 7.4.
performed by co-injection experiments. To assess the biologi-
cal relevance of the results, the rate constants of the strand
break in 0.1 M spermidine–AcOH (pH 7.4) were measured at
temperatures between 70 and 90 °C, and the value at 37 °C
was estimated to be 1.4 × 10−8 s−1 from the Arrhenius plot
(Fig. 5).
McHugh and Knowland19 used DMED, because the pKa
value of one of the protonated amino groups of this compound
(7.43) is lower than that of the other one (10.23). At pH 7.4,
about half of the molecules are singly-protonated. We expected
that the same reaction might occur when biogenic amines,
such as spermidine and spermine (Fig. 3), are used instead of
DMED. Although the lowest pKa was not close to 7.4 (for
Strand breaks in other solutions
example, that of spermine is reportedly 7.9522), the solutions Next, we tested the reaction using ethylamine, which is a
of spermidine and spermine were adjusted to pH 7.4 with simple primary amine. A solution of 0.1 M ethylammonium
AcOH, and the 30-mer was heated in each solution at 90 °C for acetate (pH 7.4) was prepared, and the 30-mer containing the
6 h. The results of the HPLC analyses of the reaction mixtures (6–4) photoproduct was treated with this solution at 90 °C for
are shown in Fig. 4B and 4C. Although the product yields were 6 h. As shown in Fig. 6A, two peaks were detected at the reten-
slightly lower, the elution profiles were similar to that obtained tion time of the 5′ fragment, while the peak of the 3′ fragment
for the reaction using DMED, and the identification was looked the same as that shown in Fig. 4A. A co-injection
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 Arrhenius plot for the strand-break reaction of the (6–4) 30-mer in
0.1 M spermidine–AcOH (pH 7.4).
(6–4) photoproduct was caused by the thermal instability of
the glycosidic bond of the 3′ component of this lesion, but not
by the reaction with the amines.
Dewar valence isomer
A
chemically-synthesized 30-mer, containing the Dewar
valence isomer of the (6–4) photoproduct in the same
sequence as above,23 was treated with 0.1 M DMED–AcOH (pH
7.4) and 0.1 M sodium phosphate (pH 7.0) at 90 °C for 6 h. As
shown in Fig. 7, only very small peaks of the 3′ fragment were
detected, and the 5′ fragment was hardly identifiable. These
results showed that the glycosidic bond of the Dewar isomer is
thermally stable, although it is more alkali-labile than that of
the (6–4) photoproduct.14,18
Since this finding is important, we further investigated
whether the same holds true for the photoproducts formed
at the thymine–cytosine sequence, which is the major (6–4)
photoproduct site.10 A tetramer, d(ATCG), was irradiated with
254 nm light to form the (6–4) photoproduct, and after HPLC
purification, the (6–4) photoproduct was converted to its
Dewar valence isomer by irradiation with Pyrex-filtered light
from a high-pressure mercury lamp. The glycosidic bond of
the Dewar isomer was obviously more labile in alkaline solu-
tions than that of the (6–4) photoproduct (Fig. 8A), in the
same manner as the photoproducts formed at TT shown in
our previous study.14 However, the stability against heat degra-
dation at pH 7.0 was the opposite, as shown in Fig. 8B and 8C,
and the results obtained for the 30-mers containing the TT
photoproducts (Fig. 6C and 7B) were reproduced.
Fig. 4 HPLC analysis of the solutions containing the (6–4) 30-mer in 0.1 M
DMED–AcOH (pH 7.4) (A), in 0.1 M spermidine–AcOH (pH 7.4) (B), and in 0.1 M
spermine–AcOH (pH 7.4) (C), after heating at 90 °C for 6 h. The insets are mag-
nifications of the product peaks. The vertical scales of the chromatograms moni-
tored at 317 nm, shown in red, are magnified by a factor of 10.
experiment revealed that one of the two smaller peaks and the
larger product peak coincided with the 5′ and 3′ fragments
obtained in the experiment using DMED, respectively. Finally,
we tried to confirm whether amines were required for the
strand break at the (6–4) photoproduct. The same oligonucleo-
tide was treated with 0.1 M ammonium acetate (pH 7.4) and
0.1 M sodium phosphate (pH 7.0) in an identical manner.
Contrary to our expectations, the same 3′ fragment produced
by the strand break at the (6–4) photoproduct was detected
even in the sample heated in the phosphate buffer, although
these treatments yielded further additional peaks assigned to
the 5′ fragment from the UV absorption spectra (Fig. 6B
and 6C). The results indicated that the strand break at the
Enzymatic removal of the 3′ blocking end
When this type of strand break occurs in cells, the 3′ end
bearing the cross-linked base and the phosphate is produced
at the gap, and this nonconventional structure blocks the
chain elongation by DNA polymerase. We investigated whether
this 3′-blocking residue could be removed enzymatically.
The excision repair cross-complementation group 1 (ERCC1)–
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 7 HPLC analysis of the solutions containing the Dewar 30-mer in 0.1 M
DMED–AcOH (pH 7.4) (A) and in 0.1 M sodium phosphate (pH 7.0) (B), after
heating at 90 °C for 6 h.
original report),28 was also used. As shown in Fig. 9B, several
nucleotides at the 3′ end of the 32P-labeled fragment were
removed in the presence of RPA, in a similar manner to the 3′-
phosphoglycolate substrate.24,25 The major products were
d(CTCGTCAG) and d(CTCGTCA), and the product bands were
not detected when the mutant XPF protein was used. The
dependence of the endonuclease activity on the RPA concen-
tration was confirmed, as shown in Fig. 9C.
Fig. 6 HPLC analysis of the solutions containing the (6–4) 30-mer in 0.1 M
ethylammonium acetate (pH 7.4) (A), in 0.1 M ammonium acetate (pH 7.4) (B),
and in 0.1 M sodium phosphate (pH 7.0) (C), after heating at 90 °C for 6 h. The
insets are magnifications of the product peaks. The vertical scales of the chroma-
tograms monitored at 317 nm, shown in red, are magnified by a factor of 10.
Discussion
The (6–4) photoproduct formed in DNA by UV irradiation is
known to be alkali-labile. Strand breaks occur when UV-irra-
diated DNA is treated with hot alkali, and we previously eluci-
dated the degradation mechanisms.13,14 The first reaction is
the hydrolytic ring opening of the 5′ base, and subsequently,
the glycosidic bond at the 3′ component is cleaved to yield an
abasic site, which leads to the strand breakage by β,δ-elimina-
tion. These two reactions occur independently, but the stab-
ilities of the bonds are different. Since extremely alkaline
conditions, such as 1 M piperidine or 0.1 M KOH, and a very
high temperature, e.g. 90 °C for a 30 min treatment, are
required for the strand break,9,29,30 these reactions are unlikely
to occur in the cells of living organisms. However, this method
is often utilized for experiments to determine the photopro-
duct sites in UV-irradiated DNA.8–12 On the other hand,
xeroderma pigmentosum complementation group F (XPF)
complex, which is an essential endonuclease for NER of UV-
damaged DNA, was chosen because it reportedly removes 3′-
phosphoglycolate,24,25 another type of 3′-blocking end. After
32P-labeling at the 5′ end of the 5′ fragment obtained in the
experiment shown in Fig. 4A, a duplex with a gap shown in
Fig. 9A was prepared as the substrate. This duplex was incu-
bated with the recombinant ERCC1–XPF complex,26 and the
product was analyzed by polyacrylamide gel electrophoresis
(PAGE). Since replication protein A (RPA) reportedly stimulates
the endonuclease activity of the ERCC1–XPF complex,27 its
effect was tested, and the D731A XPF mutant, which lacks the
nuclease activity (this mutant was referred to as D720A in the
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 9 Analysis of the ERCC1–XPF reaction. (A) The structure of the substrate. X
represents the (6–4) photoproduct without the 3’ sugar moiety. (B) PAGE ana-
lysis of the products. M represents the D731A mutant of the XPF protein.
(C) Dependence of the nuclease activity of the ERCC1–XPF complex on the RPA
concentration.
treated with 0.1 M DMED–AcOH (pH 7.4) at 90 °C for 6 h
(Fig. 2). However, the 5′ fragment was not identical to that pro-
duced by the alkali treatment in our previous study.14 Analyses
of this product by NMR spectroscopy and mass spectrometry
revealed that the structure of the (6–4) photoproduct was intact
after the strand break, while the 5′ base underwent hydrolytic
ring opening by hot alkali.13,14 The chemical structures of
these products are shown in Fig. 1. A similar strand break
occurred in a 30-mer oligonucleotide, apparently with higher
efficiency (Fig. 4A), and the same products were detected when
biogenic amines, i.e. spermidine and spermine, were used
instead of DMED (Fig. 4B and 4C). The rate constant of this
strand breakage in 0.1 M spermidine–AcOH (pH 7.4) at 37 °C
was determined to be 1.4 × 10−8 s−1. This value is larger than
the rate constants of cytosine deamination at 37 °C (1 × 10−10
and 7 × 10−13 s−1 in single- and double-stranded DNA, respecti-
vely)31 and depurination at 37 °C and pH 7.4 (3 × 10−11 s−1),32
which are both important types of spontaneous DNA damage.
We used a single-stranded oligonucleotide to determine the
rate constant, while DNA is usually in the double-stranded
form. However, the (6–4) photoproduct reportedly destabilizes
the duplex to a great extent,33 and the hydrogen bonding is
lost at the 3′ component of this photoproduct.34 Therefore, it
Fig. 8 Analysis of the strand breaks at the (6–4) photoproduct and its Dewar
valence isomer formed between thymine and cytosine. (A) Strand breaks in alka-
line solutions. Tetramers containing the (6–4) photoproduct (open symbols) and
the Dewar isomer (filled symbols) were incubated at 37 °C at pH 10 (triangles),
11 (squares), and 12 (circles), and the released 2’-deoxyguanosine 5’-phosphate
was quantified. (B and C) HPLC analysis of the solutions of the tetramers con-
taining the (6–4) photoproduct (B) and the Dewar isomer (C) in 0.1 M sodium
phosphate (pH 7.0), before and after heating at 90 °C for 6 h. The peak
detected at 3.5 min in B is 2’-deoxyguanosine 5’-phosphate.
McHugh and Knowland19 reported that strand breaks were
detected when UV-irradiated DNA was treated with DMED–
AcOH at pH 7.4. Although the reaction temperature was 90 °C,
we were interested in this report because of the physiological
pH value.
The first finding in this study was that a strand break,
similar to that caused by the hot alkali treatment, occurred
when a tetramer containing the (6–4) photoproduct was
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
is assumed that the strand break at the (6–4) photoproduct in
double-stranded DNA occurs at a similar rate.
Experimental
Materials
When the same oligonucleotide was treated with 0.1 M
ethylammonium acetate (pH 7.4), the 5′ fragment was split Compound 1 was prepared as described previously,36 using the
into two peaks with the same UV absorption spectrum intermediate in the synthesis of the oligonucleotide building
(Fig. 6A). We assumed that a DNA fragment containing an block of the (6–4) photoproduct (compound 2 in ref. 20). A
α,β-unsaturated aldehyde at the 3′ end, which was formed by tetramer and a 30-mer containing the (6–4) photoproduct,
β-elimination, was obtained, in addition to the β,δ-elimination d(AT(6–4)TG) and d(CTCGTCAGCATCT(6–4)TCATCATACAGT-
product that was exclusively produced by the DMED treatment. CAGTG), were synthesized using the dinucleotide-type building
The protonated form of DMED has pKa values of 10.23 block, as described previously.20 A 30-mer containing the
and 7.43. The value of one of the protonated amino groups of Dewar valence isomer of the (6–4) photoproduct, in the same
this compound is lower, due to the protonation of the sequence as above, was synthesized in a similar manner.23
other one. Therefore, about half of the molecules were singly- The ERCC1–XPF complex and RPA were prepared according to
protonated at pH 7.4, and the free amine in this species was the methods described previously.26,37
expected to abstract a proton in the elimination reaction.19
Since the pKa value of the conjugate acid of ethylamine
Amine treatment
is 10.63,35 this amine is almost completely protonated at The tetramer containing the (6–4) photoproduct (3.5 nmol)
pH 7.4, and thus the elimination reaction must proceed was dissolved in a 0.1 M solution (50 μL) of DMED–AcOH (pH
less efficiently. It was also proposed that the positive 7.4), and this mixture was heated at 90 °C for 6 h. The HPLC
ammonium group in the singly-protonated DMED might analysis was performed on a Gilson gradient-type analytical
interact with the phosphate group in DNA.19 However, based system equipped with a Waters 2996 photodiode array detec-
on our results, this interaction is not required, at least for the tor, using a Waters μBondasphere C18 5 μm 300 Å column
strand break.
To our surprise, the strand break was detected when the 10% for 20 min) in 0.1 M triethylammonium acetate (TEAA,
30-mer oligonucleotide was treated with ammonium acetate pH 7.0), at a flow rate of 1.0 mL min−1
(pH 7.4) or sodium phosphate (pH 7.0), although the 5′ For the NMR study, 1 (0.1 mmol) was dissolved in the above
(3.9 × 150 mm) with a linear gradient of acetonitrile (from 0 to
.
fragment became a mixture of several species (Fig. 6B and 6C). solution (10 mL), and the mixture was heated at 90 °C under a
In all of the reactions, the 3′ product was the same 5′-phos- reflux condenser. The reaction was monitored by reversed-
phorylated 16-mer oligonucleotide. These results suggested phase HPLC, using a GL Science Inertsil ODS-3 5 μm column
that the strand break was caused by thermal cleavage of the (4.6 × 250 mm) with a linear gradient of 0–7.5% acetonitrile in
glycosidic bond at the 3′ component of the (6–4) photoproduct. 0.1 M TEAA (pH 7.0). After 8 hours, the mixture was concen-
The amines that contained a low-pKa amino group, such as trated on a rotary evaporator equipped with a vacuum pump,
DMED, spermidine and spermine, probably facilitated the and the residue was dissolved in water (1 mL). The product (4)
δ-elimination reaction, which resulted in the formation of a was purified by HPLC, and the cation was exchanged to Na+
single-type 5′ fragment containing the structure shown in using a Bio-Rad AG 50W-X2 resin. The structure of 4 was
Fig. 3.
determined by NMR spectroscopy and mass spectrometry. The
The glycosidic bond of the Dewar valence isomer is more NMR measurement revealed that the sample was contami-
alkali-labile than that of the parent (6–4) photoproduct,14,18 nated with sodium acetate, which yielded a signal with a
but it was quite stable under the conditions used in this study chemical shift of 1.86 ppm in the 1H NMR spectrum
(Fig. 7 and 8). These two types of photoproducts cannot be dis- and signals with shifts of 184.2 and 26.1 ppm in the 13C NMR
tinguished by the hot piperidine treatment, because the DNA spectrum. In the high-resolution mass analysis, the m/z
strand is cleaved at both sites. However, only the (6–4) photo- value obtained for 4 was 449.1074 ([M + H]+; calculated for
product, but not the Dewar isomer, is expected to yield C15H22N4O10P: 449.1074).
fragment bands in a PAGE analysis after 5′-32P-labeled UV-
irradiated DNA is heated at 90 °C in 0.1 M sodium phosphate valence isomer (1 nmol) was dissolved in 50 μL of each solu-
at pH 7.0. tion described in the figure legends, and the mixtures were
The 30-mer containing the (6–4) photoproduct or its Dewar
Although the reaction rate is slow at 37 °C, the 3′ end of the heated at 90 °C for 6 h. For the HPLC analysis, a Waters
5′ fragment bearing the (6–4) photoproduct and a phosphate, μBondasphere C18 5 μm 300 Å column (3.9 × 150 mm) was
shown in Fig. 3, blocks replication when the strand break used at 60 °C, with a 2.5 to 12.5% acetonitrile gradient over
found in this study occurs in cells. We demonstrated its 20 min. The product (peak v in Fig. 4A) was purified under the
removal by the ERCC1–XPF complex (Fig. 9), and RPA had a same conditions, and was characterized by MALDI-TOF
stimulatory effect on the processing of this novel blocking end. mass spectrometry ([M − H]− m/z 4091.42; calculated for
Since the (6–4) photoproduct is efficiently recognized by the
C130H167N45O83P13: 4088.68). To determine the reaction rates,
NER proteins, it is repaired by the NER pathway in normal the 30-mer was treated with 0.1 M spermidine–AcOH (pH 7.4)
cells. However, this end processing may function in cells defec- at 90, 85, 80, 75, and 70 °C, and the product yields were
tive in the NER activities.
obtained from the peak areas, using the molecular extinction
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
coefficients calculated by the reported method.38 The rate con- polynucleotide kinase (20 units) and [γ-32P]ATP (0.93 MBq), in
stant at 37 °C was estimated from the Arrhenius plot (Fig. 5).
a buffer (10 μL) containing 50 mM Tris-HCl (pH 8.0), 10 mM
MgCl2, and 5 mM dithiothreitol, at 37 °C for 10 min. After
heating at 95 °C for 3 min, the ATP was removed by using a GE
NMR measurement of 4
1H, 13C, and 31P NMR spectra were measured at 30 °C on a Healthcare MicroSpin G-25 column. The substrate shown in
Varian Unity-INOVA 500 spectrometer. The 1H chemical shift Fig. 9A was prepared by mixing the three strands (each
was calibrated with internal HDO (4.70 ppm) in D2O. In the 20 pmol) in water (33 μL), heating the solution at 85 °C for
13C measurement, the 13C reference frequency was obtained by 2 min, and cooling it to 20 °C. This substrate (0.4 pmol) was
calculation from the 1H reference frequency, as reported pre- incubated with ERCC1–XPF (0.2 pmol) and RPA (1 pmol in
viously.14 The 31P chemical shift was calibrated with external Fig. 9B, and 0.05–3.2 pmol in Fig. 9C), in a buffer (10 μL) con-
trimethyl phosphate. Two-dimensional NMR spectra were taining 50 mM Tris-HCl (pH 8.0), 0.5 mM MgCl2, 5 mM 2-mer-
recorded on a 5 mm pulse field gradient probe for indirect captoethanol, and 0.1 mg mL−1 bovine serum albumin, at
detection. For the NOESY measurement, the mixing time was 30 °C for 90 min. A 96% formamide solution, containing
set to 700 ms.
9.6 mM ethylenediaminetetraacetic acid (EDTA), 48 μg mL−1
1H NMR (500 MHz) δ (ppm): 7.79 (s, 1H, pT-H6), 6.25 (dd, bromophenol blue, and 48 μg mL−1 xylene cyanol FF (10 μL),
J(H,H) = 5.2, 9.7 Hz, 1H, H1′), 5.06 (s, 1H, Tp-H6), 4.52 (m, 1H, was added to the mixture. After heating the solution at 95 °C
H3′), 4.05 (m, 1H, H4′), 3.75 (m, 2H, H5′), 2.18 (s, 3H, pT-CH3), for 5 min, the products were separated by electrophoresis on a
1.92 (dd, J(H,H) = 5.2, 13.7 Hz, 1H, H2′), 1.68 (s, 3H, Tp-CH3), 12.5% denaturing polyacrylamide gel, in a buffer containing
1.55 (m, 1H, H2′′). 13C NMR (125 MHz) δ (ppm): 178.2 (pT-C4), 89 mM Tris base, 89 mM boric acid, and 2 mM EDTA, at 35 W
177.1 (Tp-C4), 161.0 (pT-C2), 156.9 (Tp-C2), 149.1 (pT-C6), for 75 min. The bands were detected with a GE Healthcare
118.6 (pT-C5), 87.7 (d, J(C,P) = 4.6 Hz, C4′), 87.2 (C1′), 77.1 (d, Typhoon FLA 7000 image analyzer.
J(C,P) = 4.8 Hz, C3′), 74.9 (Tp-C5), 64.6 (C5′), 60.6 (Tp-C6), 38.2
(C2′), 28.0 (Tp-CH3), 16.3 (pT-CH3). 31P NMR (202 MHz) δ:
−0.468 ppm.
Conclusions
Photoproducts formed at the TC sequence
The results of this study indicated that the glycosidic bond of
A 40 μM aqueous solution (12.7 mL) of a tetramer, d(ATCG), the (6–4) photoproduct was heat-labile at neutral pH, and the
was placed in a Petri dish with an internal diameter of 9 cm, reaction rate of the strand breakage was higher than those
and was irradiated at a total 254 nm UV dose of 15 J cm−2 in observed for cytosine deamination and depurination. Analyses
an ice bath on a SpectroLinker XL-1500 UV crosslinker (Spec- of the chemical structure of the products revealed that the
tronics Corporation). The (6–4) photoproduct-containing tetra- hydrolytic ring opening of the 5′ base, which was found for the
mer with absorption maxima at 256 and 317 nm was purified alkali degradation, did not occur at pH 7.4. Several types of
by HPLC, using a Waters μBondasphere C18 5 μm 300 Å amines were effective in yielding the single product, probably
column (3.9 × 150 mm) with a linear gradient of acetonitrile by facilitating δ-elimination at the abasic site. On the other
(from 0 to 10% for 20 min) in 0.1 M TEAA (pH 7.0). TEAA was hand, the glycosidic bond of the Dewar valence isomer of this
removed by evaporation and repeated coevaporation with photoproduct was stable under the same conditions. The 3′-
water. Aliquots of the (6–4) photoproduct-containing tetramer blocking end formed by the strand break could be removed
(11.6 nmol) were dissolved in water (8.9 mL) and irradiated in by the ERCC1–XPF complex in the presence of RPA. These
the same dish on an ice bath with a 450 W high-pressure findings will extend our understanding of the properties of
mercury lamp (Ushio UM-452) through a Pyrex jacket for 5 h. UV-induced DNA damage.
A small amount of the starting material remaining after the
irradiation was removed by HPLC purification.
Alkali degradation of the tetramers containing the (6–4)
photoproduct and its Dewar valence isomer was analyzed by
Acknowledgements
incubating the oligonucleotides (3 nmol) in 50 mM sodium Support from the Takeda Science Foundation (for I.K.) is grate-
phosphate (pH 10, 11, and 12; 100 μL) at 37 °C, followed by fully acknowledged. This work was performed, in part, at the
HPLC analysis at intervals of 24 h. Thymine was added as an Osaka University Radioisotope Research Center.
internal standard for the quantification of 2′-deoxyguanosine
5′-phosphate produced by the strand break, and a GL Science
Inertsil ODS-3 column (4.6 × 250 mm) was used in this case.
Heat degradation of the tetramers at pH 7.0 was analyzed in
References
the same manner as the DMED treatment of d(AT(6–4)TG).
1 S. Iwai, in Modified Nucleosides in Biochemistry, Biotechno-
logy and Medicine, ed. P. Herdewijn, Wiley-VCH, Weinheim,
2008, pp. 97–131.
Enzymatic reactions
The 5′ fragment produced from the (6–4) photoproduct-
containing 30-mer (20 pmol) was incubated with T4
2 D. E. Szymkowski, C. W. Lawrence and R. D. Wood, Proc.
Natl. Acad. Sci. U. S. A., 1993, 90, 9823–9827.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
3 J. E. LeClerc, A. Borden and C. W. Lawrence, Proc. Natl. 21 J.-S. Taylor, D. S. Garrett and M. P. Cohrs, Biochemistry,
Acad. Sci. U. S. A., 1991, 88, 9685–9689. 1988, 27, 7206–7215.
4 C. A. Smith, M. Wang, N. Jiang, L. Che, X. Zhao and 22 R. J. Bergeron, J. S. McManis, W. R. Weimar,
J.-S. Taylor, Biochemistry, 1996, 35, 4146–4154.
5 H. Kamiya, S. Iwai and H. Kasai, Nucleic Acids Res., 1998,
26, 2611–2617.
K. M. Schreier, F. Gao, Q. Wu, J. Ortiz-Ocasio,
G. R. Luchetta, C. Porter and J. R. T. Vinson, J. Med. Chem.,
1995, 38, 2278–2285.
6 C. Masutani, R. Kusumoto, S. Iwai and F. Hanaoka, EMBO 23 J. Yamamoto, K. Hitomi, T. Todo and S. Iwai, Nucleic Acids
J., 2000, 19, 3100–3109. Res., 2006, 34, 4406–4415.
7 J.-H. Yoon, L. Prakash and S. Prakash, Genes Dev., 2010, 24, 24 S. N. Guzder, C. Torres-Ramos, R. E. Johnson, L. Haracska,
123–128.
L. Prakash and S. Prakash, Genes Dev., 2004, 18, 2283–
2291.
25 L. A. Fisher, L. Samson and T. Bessho, Chem. Res. Toxicol.,
2011, 24, 1876–1881.
26 I. Kuraoka, W. R. Kobertz, R. R. Ariza, M. Biggerstaff,
J. M. Essigmann and R. D. Wood, J. Biol. Chem., 2000, 275,
26632–26636.
8 D. E. Brash and W. A. Haseltine, Nature, 1982, 298, 189–
192.
9 F. Bourre, G. Renault and A. Sarasin, Nucleic Acids Res.,
1987, 15, 8861–8875.
10 G. P. Pfeifer, R. Drouin, A. D. Riggs and G. P. Holmquist,
Proc. Natl. Acad. Sci. U. S. A., 1991, 88, 1374–1378.
11 J. Jen, D. L. Mitchell, R. P. Cunningham, C. A. Smith, 27 W. L. de Laat, E. Appeldoorn, K. Sugasawa, E. Weterings,
J.-S. Taylor and J. E. Cleaver, Photochem. Photobiol., 1997,
65, 323–329.
N. G. J. Jaspers and J. H. J. Hoeijmakers, Genes Dev., 1998,
12, 2598–2609.
12 J.-H. Yoon, C.-S. Lee, T. R. O’Connor, A. Yasui and 28 J. H. Enzlin and O. D. Schärer, EMBO J., 2002, 21, 2045–
G. P. Pfeifer, J. Mol. Biol., 2000, 299, 681–693. 2053.
13 M. Higurashi, T. Ohtsuki, A. Inase, R. Kusumoto, 29 W. A. Franklin, K. M. Lo and W. A. Haseltine, J. Biol. Chem.,
C. Masutani, F. Hanaoka and S. Iwai, J. Biol. Chem., 2003,
278, 51968–51973.
14 N. Arichi, A. Inase, S. Eto, T. Mizukoshi, J. Yamamoto and
S. Iwai, Org. Biomol. Chem., 2012, 10, 2318–2325.
15 J.-S. Taylor and M. P. Cohrs, J. Am. Chem. Soc., 1987, 109,
2834–2835.
1982, 257, 13535–13543.
30 G. P. Pfeifer, R. Drouin, A. D. Riggs and G. P. Holmquist,
Mol. Cell. Biol., 1992, 12, 1798–1804.
31 L. A. Frederico, T. A. Kunkel and B. R. Shaw, Biochemistry,
1990, 29, 2532–2537.
32 T. Lindahl and B. Nyberg, Biochemistry, 1972, 11, 3610–
3618.
16 K. Haiser, B. P. Fingerhut, K. Heil, A. Glas, T. T. Herzog,
B. M. Pilles, W. J. Schreier, W. Zinth, R. de Vivie-Riedle and 33 Y. Jing, J. F.-L. Kao and J.-S. Taylor, Nucleic Acids Res., 1998,
T. Carell, Angew. Chem., Int. Ed., 2012, 51, 408–411. 26, 3845–3853.
17 B. P. Fingerhut, T. T. Herzog, G. Ryseck, K. Haiser, 34 J.-K. Kim, D. Patel and B.-S. Choi, Photochem. Photobiol.,
F. F. Graupner, K. Heil, P. Gilch, W. J. Schreier, T. Carell, 1995, 62, 44–50.
R. de Vivie-Riedle and W. Zinth, New J. Phys., 2012, 14, 35 H. K. Hall, J. Am. Chem. Soc., 1957, 79, 5441–5444.
065006.
36 J. Yamamoto, Y. Tanaka and S. Iwai, Org. Biomol. Chem.,
2009, 7, 161–166.
18 L.-S. Kan, L. Voituriez and J. Cadet, J. Photochem. Photobiol.,
B, 1992, 12, 339–357.
19 P. J. McHugh and J. Knowland, Nucleic Acids Res., 1995, 23,
1664–1670.
20 S. Iwai, M. Shimizu, H. Kamiya and E. Ohtsuka, J. Am.
Chem. Soc., 1996, 118, 7642–7643.
37 Y. Masuda, M. Suzuki, J. Piao, Y. Gu, T. Tsurimoto and
K. Kamiya, Nucleic Acids Res., 2007, 35, 6904–6916.
38 P. N. Borer, in Handbook of Biochemistry and Molecular
Biology, ed. G. D. Fasman, CRC Press, Cleveland, 3rd edn,
1975, vol. 1, p. 589.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.