Novel naphthalimide hydroperoxide photonucleases: The role of thiocyclic-Fused area and the difference in spectra, photochemistry and photobiological activity

Article abstract of DOI:10.1016/j.bmc.2003.09.026

Novel five- and six-membered thiocyclic-fused naphthalimide hydroperoxides (7, 8) as photonucleases were designed, synthesized via unusual isomerization in Pschorr cyclization and photooxygenation. The five-membered 7 was able to induce single-strand nick

Full text of DOI:10.1016/j.bmc.2003.09.026

Bioorganic & Medicinal Chemistry 11 (2003) 5427–5433
Novel Naphthalimide Hydroperoxide Photonucleases: The Role of
Thiocyclic-Fused Area and the Difference in Spectra,
Photochemistry and Photobiological Activity
Yufang Xu,a Xuhong Qian,b,* Wei Yao,a Ping Maoa and Jingnan Cuib
aShanghai Key Lab. of Chemical Biology, Institute of Pesticides Pharmaceuticals, East China University of
Science and Technology, POBox 544, 130 Meilong Road, Shanghai 200237, China
bState Key Laboratory of Fine Chemicals, Dalian University of Technology, POBox 40, 158 Zhongshan Road,
Dalian 116012, China
Received 23 July 2003; revised 17 September 2003; accepted 18 September 2003
Abstract—Novel five- and six-membered thiocyclic-fused naphthalimide hydroperoxides (7, 8) as photonucleases were designed,
synthesized via unusual isomerization in Pschorr cyclization and photooxygenation. The five-membered 7 wasable to induce single-
strand nicks in duplex DNA pH independently (7.0–8.5) at 0.5 mM under 366 nm, while the six-membered 8 could photonick the
duplex DNA pH dependently (7.0–8.5) at 5 mM under 450 nm and showed ‘time-controlled’ photo-bioactivity. Their thiocycles and
the angular conjugated plane have contributionsto their binding affinity with DNA and photocleaving efficiency.
# 2003 Elsevier Ltd. All rights reserved.
Introduction
intermediates.7,13ꢀ15 Coupled with the prospects of
readily functionalizing (on the imide nitrogen) with sui-
table groups, the naphthalene-derived imides are ideal
candidates for sequence-specific photooxidation and
cleavage of oligonucleotide polymer.15 The hydroper-
oxides of naphthalimides, such as 1 and 2, could effec-
tively generate hydroxyl radical under the irradiation of
long-wavelength UV-light, and exhibited promising
DNA photo-cleavage capabilities.6,7,16 The high effi-
ciency of these reagents may be attributed to their high
DNA-intercalating capabilitiesdue to their large con-
jugated planarities. Our previous studies16ꢀ18 revealed
that taking thio instead of oxo in naphthalimide (3, 4)
and using angular conjugated plane (5, 6)17 are useful
strategies for constructing novel photocleaving agents
and intercalatorsof DNA. Here we further explore a
series of novel DNA cleaving reagents bearing a larger
conjugated plane, thiocyclic fused naphthalene ring and
a hydroxyl radical generating group, hydroperoxide.
The former wasused asthe binding part, and the latter
asthe photonicking functionality. Upon photoirradia-
tion the isomers 7 and 8 exhibited significantly higher
cleaving abilitiesand interesting difference in biological
activitiesto the uspercoiled circular pBR322 DNA. In
this paper, we report (a) the synthesis and spectra of two
novel isomers of thiocyclic fused naphthalimide; (b) their
photochemistry and binding affinities with DNA; (c)
their highly efficient photodamage to DNA (Scheme 1).
DNA cleavage by synthetic nucleases is of great interest
in biology and bioorganic chemistry. During the past
decades, many DNA cleavage reagents have been
developed aspotential antitumor agentsor prothetic
groups for antisenseoligonucleotides.1ꢀ4 One of the
most well-known DNA cleavage species is hydroxyl
radical, and much effort hasbeen devoted to the devel-
.
opment of efficient methodsfor OH generation using
organic precursors by low-energy irradiation, such as
long-wavelength UV-light (l>350 nm) or more pre-
ferably by visible light irradiation.5ꢀ8 Up till now, most
of them showed photo-bioactivity around 350 nm, only
few in the presence of metal cation under visible light,9ꢀ11
and few exampleson inorganic or organic metal com-
poundsaswell asaroyloxy-pyridinethioneswith ‘time-
controlled’ photocleaving activity9ꢀ12 were known.
However, no organic hydroperoxide releasing hydroxy
radical with similar property was reported.
Among the exploited photochemical DNA cleavers,
naphthalimide has been shown to be a class of chromo-
phore capable of generating a multitude of reactive
*Corresponding author. Tel.: +86-21-6425-3589; fax: +86-21-6425-
2603; e-mail: xhqian@dlut.edu.cn
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.09.026

5428
Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
Scheme 1.
Results and Discussion
Synthesis and spectra
These reagents were synthesized from 4-bromo-3-nitro-
1,8-naphthalic anhydride20 asshown in Scheme 2. In the
synthesis, that diazonium compound took an unusual
Pschorr cyclization and gave two isomers 7 and 8 at the
1
ratio of 6:1 via H NMR.21 So far, a similar isomerism
22
wasonly observed in benzophenone derivatives,
while
usually, Pschorr cyclization means two rings take the
ring-closing reaction in the diazonium substituted posi-
tionsto give multinuclear aromatic hydrocarbon,s
especially for derivatives of phenanthrene.
Scheme 2. The synthetic route of heterocyclic-fused naphthalimides:
(a) PhSH, EtOH, reflux, 4 h, 51 yield; (b) SnCl2/HCl, 92.6% yield; (c)
Pschorr cyclization; (1) NaNO2, H2O–HCl–HOAc, 0–5 ꢁC, 2 h; (2)
CuSO4, HOAc, reflux, 2 h; 86% yield for B1 and A1 in the ratio of 6:1
mol/mol; (d) H2NCH2CHC(CH3)2, EtOH, 3 h, >90% yields; (e) hn,
O2, TPP/CH2Cl2, ꢀ10 to ꢀ15 ꢁC, 3.5 h, 36% yield for 7, 30% for 8.
The photooxygenation of the corresponding N-iso-
pentenyl-1,8-naphthalimidesat ꢀ10 ꢁC of A2 and B2 in
methylene chloride gave products 7 and 8 as shelf-stable
crystals, respectively. The hydroperoxides were identi-
1
fied via H NMR, EI-MS, IR and elemental analysis,
Photochemistry and DNA photocleavage
they displayed characteristic absorption peak of the O–
O bond vibration around 800 cmꢀ1 in FT-IR or FT-
Raman spectra. The confirmation of similar structures
Under the irradiation of light matched with their max-
imal absorptions, both of 7 and 8 showed obvious
hydroxyl radicals signal in ESR spectra with PBN (N-
tert-butyl-a-phenylnitrone) trapping, it isimportant for
the photocleavage profile whether hydroxyl radicals
were involved.
16
waswell established before.
It can be seen in Table 1 that, the UV–visabosrption
and fluorescence maximum of 7 and 8 are at longer
wavelengthscompared with thoes of
2. 7 and 8 are
strong electronic push–pull ICT (intramolecular charge
transfer) chromophores, 7 hastwo electron-donating
substituents on the same part of its naphthalene moiety
(sulfur at the para, phenyl at the meta position to the
same electron-withdrawing carbonyl group), while 8 has
two electron-donating substituents on the different part
of itsnaphthalene moiety (sulfur and phenyl at the para
position to two electron-withdrawing carbonyl groups,
respectively). Therefore, the maximal absorption and
fluorescence of 7 are at shorter wavelengths, those of 8
at long wavelengthsand 8 having stronger fluorescence.
The fluorescence quantum yield of 8 isapproximately
three timeshigher than that of 7. It implies a possibility
that besides dissipated as heat in the system by internal
conversion processes, the excitation energy of 7 might
be more easily transferred from singlet excited state to
cleave O–O bond compared with that of 8.
The cleavage activity of 7 was evaluated using super-
coiled circular pBR322 (form I) DNA (50 mM/base pair)
under photoirradiation with a transilluminator (366 nm,
2.3 mw/cm2) at a distance of 20 cm at 0 ꢁC for 0.5 h and
analyzed on a 1% agarose gel. DNA photocleavage
efficiency wasdefined asthe degree of the relaxation of
the supercoiled DNA.
Asspeculated above, 7 showed excellent cleavage activity
which can effectively photonick form I DNA into form II
DNA at a concentration aslow as0.5 mM (Fig. 1), while
no obviouscleavage wasobserved for 2 even at ashigh
as50 mM concentration.
The significantly enhanced cleavage activity of 7 con-
firmed the rationality of the design of the molecular struc-
tures. On the one hand, UV spectra showed the maximum

Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
Table 1. UV–vis and fluorescence spectra date for naphthalimide
derivatives
5429
Comp
UV–vis
max/nm (loge)
FL
max/nm (f)
Stoke shift
(nm)
l
l
7
8
2
A1
A2
B1
B2
379 (4.05)
455 (4.24)
332 (4.25)
381(3.85)
379 (3.52)
456 (3.78)
458 (4.40)
457 (0.045)
524 (0.12)
372 (0.002)
430 (0.062)
455 (0.057)
521 (0.21)
524 (0.11)
78
69
40
49
76
65
66
Figure 2. Fluorescence spectra before and after interaction of com-
pound 7 and ctDNA. Curves1–4 and 1 0–40 for compound 7 before and
after mixed with DNA, respectively, at concentrations: 1, 5, 10, and 20
mM. DNA concentration: 50 mM (bp).
Figure 1. Cleavage of supercoiled circular pBR322 DNA by hydro-
peroxides 2 and 7. Photoirradiation time 30 min. hn: 366 nm, DNA 50
mM/bp. Lane 1: DNA alone(no hn); Lanes2 ꢃ5: DNA and 2 at con-
centrationsof 2: 100, 50, 10, and 1 mM, respectively; lanes 6–9: DNA
and 7 at concentrationsof 7: 10, 5, 1, and 0.5 mM, respectively; lane
10: DNA alone.
absorption of 7, was379 nm (log e 4.05), close to the
photoirradiation at 366 nm. On the other hand, the
binding constants of 2 and 7 to calf thymusDNA mea-
sured using the fluorescence technique method23 (Fig. 2)
are 2.67ꢂ104 and 3.82ꢂ105
M
ꢀ1, respectively, the
higher intercalating capability of 7 hascontribution to
itsDNA cleaving capabilities.
When we firstly photoirradiated 8 and DNA with a
transilluminator (366 nm, 2.3 mw/cm2), no very obvious
Figure 3. Effect of concentrationsof compound 8 on the photo-
cleavage of DNA at 450 nm.
cleavage wasobesrved at ashigh as50
mM (data not
shown). Obviously, because of the mismatch between
the UV–visabosrption maximum of 8 and the photo-
irradiation wavelength, hydroxyl radicalscan not be
efficiently generated from 8. When we changed photo-
irradition wavelength to 450 nm (2.3 mw/cm2) through
a light filter, 8 showed effective cleaving capability at the
concentration of 5 mM (Fig. 3).
important than pyrone moiety, which tendsto maximize
stacking interaction with the adjacent base pairs. It was
also reported that anglicin which had the maximization
of the angular furan ring stacking, was favorable inter-
calation geometry.17,19 It implied that the five-mem-
bered thiocycle and angular conjugated plane of 7 might
be the main reason for its high intercalating ability.
To illuminate the difference in photocleaving ability
between isomers 7 and 8, DNA binding capability of 8
was also evaluated by using fluorescence analysis. The
apparent association constant Ka of hydroperoxides 8
was4.26 ꢂ103 Mꢀ1. The association constant Ka of 7
wasmore than 100 timesstronger than that of 8.
We also want to know the effect of photochemistry of 7
and 8 on their photocleaving activities. Under the irra-
diation of light, 7 and 8 should produce two parts:
hydroxyl and allyloxyl radicals, as the reactivity and
damage ability of hydroxyl radical to DNA should be
same, their abilities to produce hydroxyl and the allyl-
oxyl radicalswill play important role in their photo-
damage behavior.
On viewpoint of heterocyclic aromaticity and sulfur
with two lone pairs, the sulfur atom at the six-mem-
bered ring of 8 isnot conjugated with naphthalene ring,
while that at the five-membered ring of 7 is, therefore, 7
is an extended planar conjugation system, while 8 isnot.
These were also supported by molecular mechanics cal-
culation (Hyperchem 5.5). For example, the maximal
dihedral angel of peri-hydrogen atomsꢁon heterocyclic
Interestingly, upon exposure of the chloroform solution
of 7 to the scattered daylight at room temperature about
3 h, the corresponding hydroxyl derivative 9 wasgener-
ated. The characteristic absorption peak of the O–O
bond of 7 at 799 cmꢀ1 gradually disappeared, accom-
panied by the emergence of the broad absorption peak
of hydroxyl group at 3460 cmꢀ1 by FT-Raman and IR
ꢁ
ring isabout 30 for 8, while that is17 for 7. The pla-
narity and rigidity of 7 are helpful for itsintercalation to
DNA. It wasknown that, five-membered furan moiety
in psoralens as famous DNA intercalators is more

5430
Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
Scheme 3.
Figure 5. pH effect of the aqueousbuffer on the photocleavage of 8 at
450 nm. Photoirradiation time: 75 min, 8: 50 mM, hn:450nm, pH
value=7.0, 7.5, 8.0, 8.5. The conversion in percent of form I DNA to
form II DNA was monitored by the computer imaging system.
The photocleavage activity of 8 wasincreased remark-
ably with the prolongation of photoirradiation time,
which had not been seen in the experiment of 7 under
the similar condition (at 366 or 450 nm, 2.3 mw/cm2). 8
generated radicalsat a relatively linear rate (from 30 to
75 min). It wastime-controlled DNA photocleaving
reagents( Fig. 4), which were different from the other
hydroperoxide17,18 and those involving metal complexes
aswell asclasical Fenton reaction, where radicalsare
produced in a rapid burst.24 Up to now, to our knowl-
edge it wasthe firts organic hydroperoxide as‘time-
controlled’ DNA photocleaver, which could release
hydroxyl radicalswithout of metal oxidantsunder long-
wavelength irradiation. Thiscontinuousgeneration of
reactive hydroxy radical in the absence of metal oxi-
dantsprovidesa novel alternative to the famousFen-
ton-based chemistry, as one ever reported N-aroxyloxy-
2-thiopyridonesto produce aroyloxyl radical for ‘time-
controlled’ DNA photocleavage.12 8 wasmuch attrac-
tive for ‘time-controlled’ DNA cleavage studies and for
in vivo biomedical application involving ‘photodynamic
therapy’.
Figure 4. Effect of irradiation time on the photocleavage of DNA at
450 nm. 8: 50 mM, hn: 450 nm. The conversion in percent of form I
DNA to form II DNA was monitored by the computer imaging sys-
monitoring, but the similar phenomena was not found
for 8 under the same conditions.
Although it isknown that mots of organic hydroper-
oxideswere converted into the correpsonding ketones
derivativesvia an g-H abstraction mechanism under
irradiation by UV light with high energy,6,24,25 7 might
decomposed in a different way under scattered daylight.
Probably, because of O–O bond of hydroperoxyl group
having weaker bond energy rather than g-C–H bond,
the released and transferred energy from photoexcited
chromophore just meets the requirement of this photo-
chemical reaction, that is, it can only selectively cleave
O–O bond with the release of active species OH and
substituted allyloxyl radical, which consequently was
quenched in a hydrogen-donating solvent to give 9
(Scheme 3).
In addition, when 8 and DNA were put in bufferswith
different pH values, 8 was sensitive to the outer envir-
onment and it wasable to efficiently photocleave DNA
during pH 7.5–8.5 and itsphotobioactivity was
increased with the increasing pH value (Fig. 5). No pH-
dependent change in the photonicking efficiency was
found in the photocleavage of 7 to DNA under the same
conditions.
In the photocleavage profile the excitation wasat 366
nm with high energy and that at 450 nm with low
energy, so, the above experimental results at least
implied that hydroxyl radical ismore eaisly formed
from 7 than 8, which resulted in their differences in
DNA photocleaving capability.
So far ones wonder why some compounds show time-
controlled or pH-independent DNA photocleaving
activity and some others not. In our case, the isomers 7
and 8 might provide an opportunity to approach the
reasons. Their differences in photochemical stability and
availability of radicals possibly were major cause to
result in ‘time-controlled property’. 8 hashigh photo-
chemical stability and its radical was not easy to form.
However, 7 with lower photochemical stability will
easily give radicals at a rapid burst under the irradiation
of light. Similarly, the presence of the hydroxyl anion
Therefore, the difference of 7 and 8 in photobiological
property dependson that in molecular and electronic
structure of their thiocyclic fused areas. These revealed
that the high DNA intercalating capability and high
radical-producing ability of 7 have contributionsto its
efficient DNA cleaving capability, by comparison with
those of 8. However, It should be stressed that for effi-
cient photonick DNA at mM level concentration the
photosensitizing wavelength of 7 wasmuch shorter than
that of 8.

Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
5431
wasvery important for 8 to produce radicalsand high
pH value wasfavorable, while the presence of hydroxyl
anion did not give much help for the formation of radi-
calsderived from 7.
8.37 (d, J=7.6 Hz, 1H, 5-H), 8.15 (s, 1H, 2-H), 8.15 (t,
J=7.6 Hz, J0=8.5 Hz, 1H, 6-H), 7.22 (m,2H, 30-H, 50-
H), 7.15 (m, 1H, 40-H), 7.04 (m, 2H, 20-H, 60-H), 5.08 (s,
2H,–NH2). EI-MS (m/z,%): 321 (M+, 100), 277
(M+ꢀCO2, 6.16).
Benzothiophenonaphthalic anhydride (A1) and benzo-
thioxanthene-dicarboxylic anhydride (B1). Sodium
nitrite (0.7 g) and glacial acid (2 mL) were added drop-
wise to concd sulfuric acid (10 mL). The mixture was
cooled to 5 ꢁC and to it wasadded, portionwise over 1
h, 3-amino-4-phenylthio-1,8-naphthalic anhydride (3.2
g). After stirring for 1 h, the dark red viscous liquor was
added over 90 min to a boiling solution of copper sul-
fate in water and glacial acetic acid. After the addition
wascomplete, the liquor wasrefluxed for a further 30
min, cooled, filtered, dried and an orange solid (2.64 g,
86.3%) was collected. Recrystallisation from DMF gave
deep red needles, but TLC indicated it contained com-
pound A1 (Rf=0.54) and B1 (Rf=0.68), which were
separated on preparative thin layer chromatography
using dichloromethane as eluent (three times) and gave
two pure compounds. The compound A1 wasyellow-
green solid, mp 284–286 ꢁC. 1H NMR (d-DMSO,
300 MHz): d 9.4 (s, 1H, 7-H), 8.75 (m, 2H, 3-H, 1-H),
8.59 (dd, J1=1.0 Hz, J2=7.4 Hz, 1H, 11-H), 8.28 (m,
1H, 2-H), 8.04 (dd, J1=8.2 Hz, J2=7.4 Hz, 1H, 8-H),
7.68 (m, 2H, 9-H, 10-H); EIMS (m/z, %): 306 ([M+2]+,
11.4), 304 (M+, 100). The compound B1 wasorange red
Conclusion
7, with high DNA binding ability, wasvery eaisly
decomposed to unusually give hydroxyl derivative
through losing hydroxyl radical in scattered daylight
and it exhibitshighly-effective photocleaving activity
under the irradiation of light at 366 nm at 0.5 mM with
pH-independent property (pH 7.0–8.5). 8 efficiently
photo-cleaved DNA at a concentration of 5 mM, it was
able to give promising results at pH 7.5–8.5 and its
photobioactivity was increased with the increasing of
pH value. Probably 8 wasthe firts organic hydroper-
oxide which exhibitshighly effective photocleavage with
‘time-controlled’ property, itsradicalscan be generated
easily under the irradiation of visible light (450 nm). In
general, it was demonstrated that sulfur-heterocyclic
angular conjugated plane wasan effective tsrategy for
constructing novel photocleavers.
Experimental
Materials and methods
solid, mp >300 ꢁC, H NMR (-dDMSO, 300 MHz): d
1
Melting pointswere taken on a digital melting point
apparatusWRS-1 made in Shanghai and it wasuncor-
rected. Infrared spectra were recorded on a Nicolet FT
8.55(m, 3H, 2-H, 6-H, 7-H), 8.36 (d, J=8.0 Hz, 1H, 2-
H), 7.83 (d, J=8.1 Hz, 9-H), 7.56ꢃ7.70(m, 3H, 10-H,
11-H, 12-H); EIMS (m/z,%): 306 ([M+2]+, 8.7), 304
(M+, 100).
1
IR-20SX, mass spectra on a Hitachi M80, H NMR on
a Bruker AM-300 or AM-500 using TMS as an internal
standard. Combustion analysis for elemental composi-
tion wasdone on Italy MOD.1106 analyzer. Absorption
spectra were measured on Shimadzu UV-265, fluores-
cence spectra on a Hitachi 850.
N-Iisopentenyl-benzothiophenonaphthalimide (A2). Com-
pound A1 (0.9 g) wasrefluxed in ethanol (40 mL) with
isopentenylamine (0.4 g) for 3–5 h. After removal of the
solvent, the residue was dissolved in CH2Cl2 and then
purified on column chromatography with petroleum–
dichloromethane (1:1) aseluent. The greenihs-yellow
main product wascollected and recrystallized from pet-
roleum–chloroform to give greenish-yellow needles
Synthesis and photooxygenation
3-Nitro-4-phenylthio-1,8-naphthalic
anhydride.19
4-
(1.02 g), mp 207–208 ꢁC, H NMR (CDCl3, 500 Hz): d
1
Bromo-3-nitro-1,8-naphthalic anhydride (6.44 g) was
stirred under reflux in ethanol (60 mL) with thiophenol
(2.6 mL) for 5 h. The liquor wasreduced in volume to
30 mL and filtered, washed with some ethanol, dried,
giving 6.83 g (97.3%) of 3-nitro-4-phenylthio-1, 8-
naphthalic anhydride. The recrystallization from glycol
monomethyl ether gave long golden needles. Mp 177–
178 ꢁC (lit.19 178–179 ꢁC).
1.79 (s, 3H, 30-CH3), 1.97 (s, 3H, 30-CH3), 4.66 (d,
J=6.7 Hz, 2H, –NCH2–), 5.39 (t, J1=J2=6.7 Hz, 1H,
20-H), 7.32–7.37 (m, 2H, 9-H, 10-H), 7.49 (m, 1H, 2-H),
7.67 (d, J=7.6 Hz, 1H, 8-H), 7.79 (d, J=7.6 Hz, 11-H),
7.92 (d, 8.0 Hz, 1H, 1-H), 8.26 (d, J=7.1 Hz, 1H, 3-H),
8.54 (s, 1H, 7-H); EI-MS (m/z, %): 373([M+2]+, 5.4),
371(M+, 53.5), 302 ([MꢀCH2CH¼CMe2]+, 100); n
(KBr)max/cmꢀ1: 3060, 2980, 2850, 1690, 1660, 1580,
1440. Anal. calcd for C23H17NO2S: C, 74.37, H, 4.62, N,
3.77. Found: C, 73.98, H, 4.62, N, 3.70.
3-Amino-4-phenylthio-1,8-naphthalic anhydride.19 The
above nitro compound (2.7 g) wasstirred into a mixture
of stannous chloride (8.81 g) and concentrated hydro-
chloric acid (12 mL). After warming to 40 ꢁC, the tem-
perature rose spontaneously to 85 ꢁC and was
maintained at 85 ꢁC for 1 h (color change from orange
to olive-green). The suspension was cooled and filtered
to give 2.4 g crude product. Recrystallization from pyr-
idine gave greenish-yellow needles, mp 224–225 ꢁC19 of
3-amino-4-phenylthio-1,8-naphthalic anhydride. 1H
NMR (CDCl3, 300 Hz): d 8.69 (d, J=8.5 Hz, 1H, 7-H),
N-Isopentenyl-benzothioxanthenediimide (B2)
In the similar manner to the synthesis of A2, B2 was
obtained asorange red needle,s mp 202–203 ꢁC, 1H
NMR (CDCl3, 500 Hz): d 1.74 (s, 3H, 30-CH3), 1.77 s,
3H, 30-CH3), 4.73 (d, J=7.5 Hz, 2H, –NCH2–), 5.35 (t,
J1=J2=7.5 Hz, 1H, 20-H), 7.39 (m, 3H, 10-12H), 7.49
(d, J=8.0 Hz, 1H, 9-H), 8.20 (m, 2H, 1-H, 7-H), 8.42

5432
Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
(d, J=8.0 Hz, 6-H), 8.62 (d, J=8.0 Hz, 1H, 2-H); EI-
MS (m/z, %): 373 ([M+2]+, 5.1), 371 (M+, 41.3), 302
Intercalation studies of compounds 7 and 8 to DNA
([M+H–CH2CH¼CMe2]+, 100);
n
(KBr)max/cmꢀ1
:
0.1 mL of solution of a compound in DMSO (10ꢀ3
–
3080, 2960, 2930, 1690, 1645, 1585, 1380. Anal. calcd
for C23H17NO2S: C, 74.37, H, 4.62, N, 3.77. Found: C,
74.08, H, 4.89, N, 3.79.
10ꢀ4 M) mixed with 0.1 M Tris–HCl buffer (pH 7.4) to
10 mL. Then, two groupsof sampleswere prepared in
the concentration of chemical at 10ꢀ5–10ꢀ6 M; one
contained calf thymusDNA 50 mM, the other contained
no DNA but had the same concentration of chemical as
control. All the above solution was shaken for 3 days at
25 ꢁC in the dark. Fluorescence wavelengths and inten-
sity areas of samples were measured at following condi-
tions: excitation: 365 or 450 nm, emission: 380–520 or
470–650 nm.
Benzothiophenonaphthalimide hydroperoxide (7). A solu-
tion of A2 (200 mg) and tetraphenylporphyrin (TPP, 5
mg) in 40 mL of CH2Cl2 wasirradiated externally by
meansof a sodium lamp (150 W) at ꢀ10 to ꢀ20 ꢁC for
3–4 h while passing a continuous slow stream of dry
oxygen gas through the solution. The solution was con-
centrated and purified by preparative thin layer chro-
matography on silica gel using petroleum ether–acetate
(2:1) aseluent. The zone with Rf=0.6 wascollected as
product and gave greenish-yellow solid (36%). Mp 174–
Photocleavage of supercoiled DNA using compounds 7
and 8
175 ꢁC, H NMR (CDCl3, 500 Hz): d 2.01 (s, 3H, 30-
300 ng pBR 322DNA (form I), 1 mL of solution of
chemical in DMSO and 10 mM Tris–HCl buffer (pH
7.6) were mixed to 10 mL and stood for 10 min at 0 ꢁC,
then irradiated for 30 min or more with light (2.3 mW/
cm2, 365 or 450 nm) using lamp placed at 20 cm from
sample. Supercoiled DNA runs at position I, nicked
DNA at position II, and linear DNA at position III.
The samples were analyzed by gel electrophoresis in
1% agarose and gel was stained with ethidium
bromide.
1
CH3), 4.70 (m, 3H, N-CH2-CH-,), 5.12 (d, J=1.4 Hz,
2H, 30-CH2=), 7.56 (m, 2H, 9-H, 10-H), 7.75 (m, 1H, 2-
H), 7.92 (d, J=7.6 Hz, 1H, 8-H), 8.19 (d, J=7.6 Hz,
1H, 11-H), 8.32 (d, J=8.0 Hz, 1H, 1-H), 8.53 (d, J=7.1
Hz, 1H, 3-H), 9.04 (s, 1H, 7-H), 10.17 (br, 1H, -OOH);
EI–MS (m/z, %): 386 ([MꢀOH]+, 5.7); n (KBr)max
/
cmꢀ1: 3258, 1688, 1640, 1587, 1335, 906, 799, 780; FT-
Raman: 3065, 1689, 1587, 1400, 1377, 799 cmꢀ1. HRMS:
calcd for C23H17NO4S: 403.4878; Found: 403.0874.
Benzothiophenonaphthalimide hydroperoxide (8). In the
similar manner to the synthesis of 7, 8 wasobtained as
Acknowledgements
orange red solid. Mp 210–211 ꢁC, H NMR (DMSO-d,
1
500 Hz): d 1.79 (s, 3H, 30-CH3), 4.07 (m, 1H, 10-CH2),
4.32 (m, 1H, 10-CH2), 4.64 (m, 1H, 20-CH), 4.88 (2s, 2H,
CH2¼), 7.53 (m, 2H, 10-H, 11-H), 7.62 (dd, J1=1.6 Hz,
J2=7.6 Hz, 1H, 9-H), 7.78 (d, J=8.0 Hz, 1H, 12-H),
8.35 (d, J=8.0 Hz, 1H, 1-H), 8.48–8.55 (m, 3H, 2-H, 6-
H, 7-H), 10.59 (br, 1H,-OOH); EI-MS (m/z, %): 386
([MꢀOH]+, 20.4); n (KBr)max/cmꢀ1: 3250, 2940, 1688,
1635, 1587, 1385, 800, 770 cmꢀ1. HRMS: calcd for
C23H17NO4S: 403.4878; Found: 403.0875.
Thiswork wassupported by the Fuk Ying Tung Foun-
dation, The Ministry of Education of China, the
National Natural Science Foundation of China, Shang-
hai Committee of Science and Technology and Shanghai
Committee of Education.
References and Notes
1. Chatterjee, M.; Rokita, S. J. Am. Chem. Soc. 1991, 113,
5116.
2. Saito, I.; Takayama, M.; Sugiyama, H.; Nakamura, T. In
DNA and RNA Cleavers and Chemotherapy of Cancer and
Viral Diseases; Menunier, B., Ed.; Kluwer: Netherlands, 1996;
pp 163.
3. Breslin, D. T.; Coury, J. E.; Anderson, J. R.; McFail-Isom,
L.; Kan, Y.Z.; Williams, L.D.; Bottomley, L.A; Schuster, G.B.
J. Am. Chem. Soc. 1997, 119, 5043.
4. (a) Armitage, B. Chem. Rev. 1998, 98, 1171. (b) Pogozelski,
W. K.; Tullius, T. D. Chem. Rev. 1998, 98, 1089.
5. Bielski, B. H.; Gehicki, J. M. Free Radic. Biol. 1977, 1, 1.
6. Saito, I. Pure Appl. Chem. 1992, 64, 1305.
7. (a) Saito, I.; Takayama, M.; Sugiyama, H.; Nakatani, K.;
Tsuchida, A.; Yamamoto, M. J. Am. Chem. Soc. 1995, 117,
6406. (b) Matsugo, S.; Saito, I. Tetrahedron Lett. 1991, 25,
2949. (c) Saito, I.; Takayama, M.; Kawanishi, S. J. Am. Chem.
Soc. 1995, 117, 5590. (d) Saito, I.; Takayama, M.; Sakurai, T.
J. Am. Chem. Soc. 1994, 116, 2653. (e) Saito, I.; Takayama,
M.; Matsuura, T. J. Am. Chem. Soc. 1990, 112, 883.
8. (a) Adam, W.; Berger, M.; Cadet, J.; Dall’Acqua, F.; Epe,
N-hydroxyl-isopentenyl benzothiophenonaphthalimide (9).
Upon exposure of the chloroform solution of 7 to the
scattered day light at room temperature for about 3 h,
then collected sample at Rf=0.4 on preparative TLC
with methylene chloride aseluent, after removal of sol-
vent yellow solid was obtained. Mp 95–97 ꢁC, 1H NMR
(CDCl3, 500 Hz): d 1.96 (s, 3H), 4.42 (d, J=5.9 Hz,
2H), 4.53 (t, J=5.9 Hz, 1H), 4.98 (s, 1H), 5.18 (s, 1H),
7.56 (m, 2H), 7.78 (m, 1H), 7.94 (m, 1H), 8.23 (m, 1H),
8.35 (d, J=7.8 Hz, 1H), 8.56 (d, J=7.3 Hz, 1H), 9.11 (s,
1H), 9.78 (br, 1H); EI-MS m/z 387 (M+); IR (KBr):
3458, 1699, 1652, 1587, 1334, 913, 780 cmꢀ1; FT-
Raman: 3061, 1700, 1588, 1401, 1377 cmꢀ1. HRMS:
calcd for C23H17NO3S, 387.0929; Found: 387.0923.
Spectroscopic measurements
The compounds were dissolved in absolute ethanol to
give 10ꢀ5 M solutions which were read with Shimadzu
UV-265 for absorption spectra and with Perkin-Elmer LS
50 using quinine sulphate in sulphuric acid as quantum
yield for fluorescence spectra.
B.; Frank, S.; Ramaiah, D.; Raoul, S.; Saha-Moller, C. R.;
¨
Vedaldi, D. Photochemi. Photobiol. 1996, 63, 768. (b) Adam,
W.; Groer, P.; Mielke, K.; Saha-Moeller, C. R.; Kiefer, W.;
Nagel, V.; Schneider, F. W.; Ballmaier, D.; Schleger, Y.; Epe,

Y. Xu et al. / Bioorg. Med. Chem. 11 (2003) 5427–5433
5433
B. Photochem. Photobil. 1997, 66, 26. (c) Adam, W.; Cadet, J.;
Dall’Acqua, F.; Epe, B.; Ramaiah, D.; Saha-Moeller, C. R.
Angew. Chem., Int. Ed. Engl 1995, 34, 107.
9. Sarker, A. K.; Tabata, M.; Wasaaki, K. Anal. Sci. 1997, 13,
509.
17. Qian, X.; Huang, T.; Wei, D.; Zhu, D.; Fan, M.; Yao, W.
J. Chem. Soc., Perkin Trans. 2 2000, 715.
18. (a) Qian, X.; Mao, P.; Yao, W.; Guo, X. Tetrahedron Lett.
2001, 43, 2995. (b) Yao, W.; Qian, X.; Hu, Q. Tetrahedron
Lett. 2000, 41, 7711. (c) Yao, W.; Qian, X. J. Fluorine Chem.
2000, 106, 69. (d) Yao, W.; Qian, X. Dyes Pigments 2000, 106,
69.
10. Kumar, C. V.; Tan, W. B.; Bett, P. W. J. Inorg. Chem.
1997, 68, 177.
11. Chatterjee, S. R.; Srivastava, T. S.; Kamat, J. P.; Devasa-
gayam, T. P. A. J. Porphyrins Phthalocyanines 1998, 2, 337.
12. Theodorakis, E. A.; Wilcoxen, K. M. J. Chem. Soc.,
Chem. Commun. 1996, 1927.
13. Aveline, B. M.; Matsugo, S.; Redmond, R. W. J. Am.
Chem. Soc. 1997, 119, 11785.
19. Demaret, J. P.; Brunie, S. J. Photochem. Photobiol. B
1990, 6, 207.
20. (a) Kadhim, A. M.; Peters, A. T. J.S.D.C 1974, 90, 153.
(b) Peters, A. T.; Bide, M. J. Dyes Pigments 1985, 6, 267. (c)
Grayshan, P. H.; Kashim, A. M.; Peters, A. T. J. Heterocyclic
Chem. 1974, 11, 33.
14. Matsugo, S.; Kawanishi, S.; Yamamoto, K.; Sugiyama,
H.; Matsuura, T.; Saito, I. Angew. Chem., Int. Ed. Engl. 1991,
30, 1351.
15. (a) Rogers, J. E.; Kelly, L. A. J. Am. Chem. Soc. 1999,
121, 3854. (b) Rogers, J. E.; Weiss, S. J.; Kelly, L. A. J. Am.
Chem. Soc. 2000, 122, 427.
16. (a) Tao, Z.; Qian, X.; Wei, D. Dyes Pigments 1996, 31,
245. (b) Tao, Z.; Qian, X. Dye Pigments 1999, 43, 139. (c) Tao,
Z.; Qian, X.; Fan, M. Bull. Chem. Soc. Jpn. 1999, 72, 1571.
21. Qian, X.; Cui, J.; Zhang, R. Chem. Commun. 2001, 2656.
22. Karady, S.; Abramson, N. L.; Dolling, U. H.; Douglas,
A. W.; McManemin, G. J.; Marcune, B. J. Am. Chem. Soc.
1995, 117, 5425.
23. Gupta, M.; Ali, R. J. Biochem. 1984, 95, 12533.
24. Howard, J. A.; Chenier, J. H. B. Can. J. Chem. 1980, 58,
2808.
25. Saito, I.; Takayama, M.; Matsuura, T.; Matsugo, S.;
Kawanishi, S. J. Am. Chem. Soc. 1990, 112, 883.