Thio-heterocylic naphthalimides with aminoalkyl side chains: Novel alternative tools for photodegradation of genomic DNA without impairment on bioactivities of proteins

Article abstract of DOI:10.1016/j.bmc.2004.12.012

Thio-heterocylic naphthalimides (R1-R5) were designed, synthesized and evaluated as nonmetallic and long-wavelength photocleavers. Some of them showed highly efficient abilities in the degradation of plasmid and genomic DNA under the mild conditions witho

Full text of DOI:10.1016/j.bmc.2004.12.012

Bioorganic & Medicinal Chemistry 13 (2005) 1615–1622
Thio-heterocylic naphthalimides with aminoalkyl side chains:
novel alternative tools for photodegradation of genomic
DNA without impairment on bioactivities of proteins
a,
Qing Yang, Xuhong Qian, Jianqiang Xu, Yuanshe Sun and Yonggang Li
b,
a
a
b
*
*
a
Department of Bioscience and Biotechnology and State Key Laboratory of Fine Chemistry, Dalian University of Technology,
No 2 Lingong Road, Dalian 116024, PR China
b
Shanghai Key Laboratory of Chemical Biology, Institute of Pesticides and Pharmaceuticals, East China University of Science
and Technology, No 130 Meilong Road, Shanghai 200237, PR China
Received 27 October 2004; revised 8December 2004; accepted 8December 2004
Available online 12 January 2005
Abstract—Thio-heterocylic naphthalimides (R1–R5) were designed, synthesized and evaluated as nonmetallic and long-wavelength
photocleavers. Some of them showed highly efficient abilities in the degradation of plasmid and genomic DNA under the mild con-
ditions without obvious impairment on the proteinsꢀ bioactivities, when compared with frequently applied nucleic acids removal
reagents or precipitants. Their differences in photodegradation selectivity to DNA rather than proteins were dependent on their
photodamage mechanisms and binding modes with bio-macromolecules. When maize genomic DNA was used as substrate,
ꢀ
4
2
.38 · 10 M of R5 exhibited the nuclease activity of 8Unit DNase I, R5 has some characteristic as a typical catalyst as no con-
sumption after two cycles of the photodegradation for DNA. The experiments of enzymatic activity assay and immunology activity
analysis showed that R5 was safe to proteins, suggesting its potential in the removal of transgenic material during the preparation of
bioactive proteins or enzyme preparations.
ꢀ
2004 Elsevier Ltd. All rights reserved.
4
5
1
. Introduction
sulfate, and protamine sulfate. ). All of these are expen-
sive and possibly toxic, and also they may complex unde-
sirably with certain enzymes resulting in a loss of purity.
Therefore, it is very important that these reagents are
not harmful to the proteins, at least without obvious
impairment on proteinsꢀ bioactivities and that they can
easily be removed in the purification procedure.
The removal of nucleic acids from protein or enzyme
preparations, especially the removal of transgenic mate-
rials from biochemical preparations is of biotechnolog-
ical and commercial importance due to the big issue of
transgenic safety among societies. Normally, nucleic
acids were removed by precipitation, or degraded by
the addition of exogenous nucleases.
1
Nucleic acid degrading enzymes are the often choice
early in the purification process. However, besides
expensive, majority of natural enzymes show an obligate
requirement of metal ions for their activities, which pre-
vent their uses in the presence of metal chelators or che-
Ammonium sulfate precipitation can be effective in
removing nucleic acids from the mixture of proteins,
however, it will remove some proteins at the same time.
Nucleic acids bind efficiently to anion exchangers (DEAE-
and Q-) but sometimes traces of nucleic acids can escape
and be mixed with the protein of interest. Various more
specific precipitants have been used (e.g., polyethyleneim-
6
–8
lating buffers. Therefore, novel alternatives for
nucleic acids degradation are of great potential.
Chemical or photochemical reaction reagents as artifi-
cial nucleases have received considerable attentions for
their potential applications
2
3
ine, cetyltrimethylammonium bromide, streptomycin
9
–11
in molecular biology,
since they can be easily designed and synthesized. How-
ever, less attention was paid to their applications for the
separation of biomolecules in biotechnology. The rea-
sons may be that they could not avoid (1) the obligate
Keywords: Enzyme; Nucleic acids; Photocleaver.
*
Corresponding authors. Tel.: +86 411 3673466; fax: +86 411
673488; e-mail addresses: qingyang@dlut.edu.cn; xhqian@ecust.
edu.cn
3
0
968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.12.012

1616
Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
1
2–17
requirement of metal ions,
and (2) the impairment
2-(3-dimethylamino-propyl)-6-thia-2-aza-benzo[def]chry-
1
4
on proteins. To the best of our knowledge, the latter
reason should be the most prominent limitation of their
applications. Kumar and co-workers ever applied orga-
nometallic chemicals as proteases for photocleavage of
sene-1,3-dione, showed the best results in highly-efficient
photocleavage for genomic DNA, and its impairment
effect on proteinsꢀ bioactivities was not observed by
compared with frequently applied nucleic acids removal
reagents or precipitants.
1
8,19
hen egg lysozyme and bovine serum albumin
through cation radical and free-radical mechanism, indi-
cating photoactivated compounds easily caused damage
to proteins. Similarly, we ever observed that H O and
2. Results
2
2
other organic peroxides could form hydroxyl free radical
to efficiently degrade genomic DNA in the presence of
2
.1. Syntheses and spectra of compounds
2
+
Fe , while they damaged proteins at the same time if
prolonging exposure time. The object for this paper is
to explore a synthetic DNA photocleaver as a kind of
efficient but inexpensive nuclease alternative for the deg-
radation of genomic DNA, which should exhibit nucle-
ase-similar activities and does not have obvious
impairment on proteinsꢀ bioactivities under physiologi-
cal conditions and in the absence of metal ions.
The 2-substituted-6-thia-2-aza-benzo[def]chrysene-1,3-
diones (R1–R3 and R5) were synthesized according to
2
0b
the literature
Scheme 2), while 5-butyl-10-thia-5,8,9-triaza-cyclo-
penta[a]phenalene-4,6-dione (R4) was synthesized
from 4-bromonaphthalic anhydride
(
2
0c
according to the literature from 4-bromo-3-nitronap-
thalic anhydride (Scheme 3). Separation and purification
of products were performed through a silica gel column
chromatography. All the structural details were con-
Many naphthalimide derivatives were well known for
their wide uses as anticancer drugs or photonucleases.
We ever previously reported that the presence of thio-
moiety on chromophore would obviously promote the
1
firmed by IR, H NMR, MS, and elemental analysis.
The comparisons of compoundsꢀ spectral properties
suggested that there might be some differences in their
photochemical or photophysical behaviors. 2-Substi-
tuted-6-thia-2-aza-benzo[def]chrysene-1,3-diones (R1–
R3 and R5) showing the emission at 520 nm with strong
fluorescence and absorptions around 464 nm with the
similar intensities, were composed of naphthalimide
chromophore and electron-donating benzenethioxyl
group, which implied that the electron transfer from
2
0
photocleavage. Meanwhile, N-alkyl substituted amino-
alkyl side chain, which commonly appeared in antican-
2
1
cer drugs,
could possibly enhance the affinity
interactions of the parent structure with DNA. Thus,
the design, synthesis and evaluation of thio-heterocyclic
naphthalimides R1–R5 with N-alkyl substituted amino-
alkyl side chain (Scheme 1) became our target.
2
0b
them might easily occur during DNA photocleavage.
Their structures were synthesized and confirmed by IR,
H NMR, MS, and elemental analysis. Under physio-
logical conditions, R2, R4, and R5, photoactivated at
1
5
4
- Butyl-10-thia-5,8,9-triaza-cyclopenta[a]phenalene-
,6-dione (R4), composed of naphthalimide chromo-
phore and thiadiazole, which would produce triplet state
or several active radical intermediate photochemi-
400 nm, showed high cleavage activities for plasmid
and genomic DNA in the absence of metal ions. R5,
2
0c
cally, had absorption at 362 nm, its weak fluorescence
U < 0.0003) implied that their excitation energy might
be transferred from the excited singlet state to triplet
(
R
N
CH CH CH CH
state to show some characters of radicals in DNA
2
photocleavage (Table 1).
2
2
2
3
0c
O
N
O
R1
R2
R= CH CH CH CH
2 2 2 3
O
O
2
.2. Photocleavage activity of plasmid DNA
R= CH2CH2N(CH3)2
For artificial photonuclease or photoprotease, light at
wavelengths longer than 300 nm is a highly selective
cofactor and the number of side reactions is limited be-
cause nucleic acids and most proteins will be transparent
under the light except the chromophore of the photocle-
R3 R=CH2CH2N
NH
N
N
S
S
R=
CH CH CH N(CH )
3 2
R5
2
2
2
R4
9
,18
Scheme 1. Structural features of DNA photocleavers.
aver.
The investigation showed the above compounds
R
O
O
O
O
O
O
O
O
O
O
N
O
b
c
a
Br
S
S
S
NH2
Scheme 2. Syntheses of R1, R2, R3, and R5. Reagents and conditions: (a) ortho-aminobenzenethiol, K
HOAc, NaNO , H O, HOAc, reflux, 1.5 h; (c) amine, ethanol, reflux, 2 h.
, 0–5 ꢁC, 20 min, CuSO
2
3 2
CO , DMF, reflux, 30 min; (b) H O, HCl,
2
4
2

Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
1617
CH CH CH CH
2
2
2
3
O
O
O
O
O
O
O
O
O
O
O
O
O
N
O
a
b
d
c
NO2
NO2
NH2
N
S N
N
S N
R-4
Ph
S
Ph
S
Br
Scheme 3. Synthesis of compound R4. Reagents and conditions: (a) benzyl mercaptan, K
2
2
CO
3
, DMF, Ar, 80 ꢁC, 8h; (b) SnCl
2
Æ2H
2
O, HCl, 90 ꢁC,
h; (c) HOAc, H
2
O, HCl, NaNO
2
, 0 ꢁC, 50 min; (d) butylamine, EtOH, reflux, 2 h.
a,b
Table 1. Spectra data of compounds R1–R5
the reaction very efficiently. It was also observed that
SOD (superoxide anion radical killer) accelerated the
rate of DNA-cleaving reaction, because the hydrogen
peroxide produced by SOD from superoxide anion rad-
ical would decompose to produce more active hydroxyl
radicals in the presence of UV light or trace of metal
ions. Thus, we presumed that electron transfer involved
the formation of superoxide anion radical played an
Compound
UV kmax/nm (lg e)
FL kmax/nm (/)
R1
R2
R3
R4
R5
462 (4.45)
462 (4.42)
461 (4.20)
362 (3.87)
462 (4.37)
519 (0.89)
521 (0.51)
521 (0.47)
470.4 (<0.0003)
521 (0.63)
a
In absolute ethanol.
With fluorescein as standard (/ = 0.90).
2
0b
b
important role in the photodamage,
that is, under
the irradiation of light, the thio-heterocyclic aromatic
rings of R1–R3, and R5 took an electron from DNA
and their imide moieties also transfer an electron to oxy-
gen for the formation superoxide anion radical, which at
could photocleaved plasmid DNA under the irradiation
of light at long-wavelength more than 400 nm, which is
safe for the operator and control of side reaction.
2
0b
last damaged DNA. As to R4 (10 lM), the investiga-
tion showed that ethanol inhibited its reactivity a little
while DTT did a lot. Thus, we presumed that free radi-
cals mechanism (partly) as well as the similar electron
transfer mechanism (mainly) played roles in the photo-
R1–R5 (5 lM) dissolved in DMSO could converse
supercoiled pBR322 DNA (form I) buffered in 20 mM
Tris–HCl, pH 7.5, into relaxed circular DNA (form II)
at different conversions and followed to linear DNA
2
0c
damage, the triplet state of triazole was possibly in-
0c
volved for this free radical mechanism.
2
(
form III) with a transluminator (400 nm) at a distance
of 20 cm at 25 ꢁC for 2 h photoactivation (Fig. 1). R4
and R5 showed high photocleaving efficiencies.
2.3. Photodegradation activity of genomic DNA
In our case, all the cleavage was observed without using
piperidine, although ones expected that the excited
naphthalimide might be able to directly oxidize the
guanine in DNA and the oxidation would be detected
So far, the evaluations for most of the known DNA
photocleavers were localized to plasmid DNA or synthe-
sized oligonucleotides, which could not reflect the real
features of genomic DNA. Therefore, we evaluated the
photodegradation activities of the above compounds
for genomic DNA.
9
as strand cleavage only after piperidine treatment.
For R1–R3, and R5 (10 lM), mechanistic experiment
was performed with the addition of histidine (6 mM),
dithiothreitol (DTT, 30 mM), superoxide dismutase
R1–R5 (0.01 M) dissolved in DMSO for 40 min photo-
activation resulted in 100% degradation efficiency when
the natural substrate, genomic DNA of maize in 20 mM
Tris–HCl, pH 7.5, was applied (Fig. 2a). It showed that
the size of degraded genomic DNA fragments could be
even smaller (<300 bp) if prolong photoactivation.
(
SOD, 100 lg/mL), and ethanol (1.7 M). It was found
that histidine (singlet oxygen quencher) had no effect
on cleavage, ethanol (free radical quencher) strangely
and slightly accelerated the reaction somewhat. How-
ever, DTT (superoxide anion radical scavenger) retarded
2
2
According to Kunitz, the cleavage of polymerized
DNA would result in an increase in absorbance at
2
60 nm. Therefore, the slope of the curves in Figure 3
represented the increased rate of the absorbance at
60 nm. The slope of 1 Unit DNase I was used as stan-
2
dard. The order of their photocleavages to genomic
DNA is R5 > R4 > R2. Under the same conditions,
ꢀ
4
2
.38 · 10 M of R5 exhibited the bioactivity of 8Unit
DNase I, suggesting it was an efficient agent for the deg-
radation of DNA (Fig. 2b). In fact, few synthetic and
nonmetallic compounds were reported to degrade a
genomic DNA, though there were many reports about
their cleavage activities to plasmid supercoiled DNA
and some oligonucleic acids.
Figure 1. 1% Agarose gel eletrophoresis assay of plasmid pBR322
DNA cleaved by artificial compounds. Lane 1, plasmid pBR322 DNA;
lane 2-5, pBR322 DNA photocleaved by compound R2, R3, R4, R5,
respectively; lane 6, R1.

1618
Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
Figure 2. (a) 2.5% Agarose gel electrophoresis assay for the photodegradation of genomic DNA in 20 mM Tris–HCl, pH 7.5, treated by 0.001 M of
R5 dissolved in DMSO at different time intervals. Sample (1 lL) was applied. Lane 1–5, 0, 20, 40, 60, 80 min; lane 6, marker UX174; (b) nuclease
activity comparison of R2, R4, R5, and DNase I. 5 lL 10 mM of the compound in DMSO into a reaction mixture containing 500 ng freshly prepared
maize genomic DNA in 15 lL 20 mM of Tris–HCl, pH 7.5. The reaction initiated with light at 400 nm and stopped at different time points. Nuclease
activity determined at 260 nm. When 3 lL DNase I (70 U/lL) from bovine pancreas applied as standard, the reaction buffer was changed to 20 mM
2
of Tris–HCl, pH 7.5, containing 4 mM of MgCl and 1 mM of dithiothreitol.
1
0
8
6
4
2
0
2
(
a)
(b)
1
0
.5
1
6.2
3
6.3
4
2
.6
2
.3
.5
0
0
4
60
510
560
610
660
blank
1
2
Wavelength (nm)
Figure 3. (a) Fluorescence spectral analysis of R5 (0.01 mM) in 20 mM Tris–HCl, pH 7.5, before and after photoactivation in the presence of maize
genomic DNA at various concentrations. Emission at 470 nm and adsorption at 522 nm; (b) decrease rate of fluorescence intensity of the compound
before and after photoactivation in the presence of different concentration of maize genomic DNA. Legends: (1) with addition of 500 ng/lL DNA, no
light; (2) with addition of 500 ng/lL, with photoactivation; (3) with another addition of 500 ng/lL DNA, no light; (4) with another addition of
500 ng/lL, with photoactivation.
2.4. Stability analysis of the compound during DNA
degradation
corresponding to the consumption during the first and
second photocleavage reaction, was only 0.363% and
0
.06%, respectively. Therefore, we concluded that R5
Only few of the known DNA photoactivated agents
actually were treated as ꢁphotonucleasesꢀ, which in an
electronically excited state react directly and catalyti-
cally with the nucleic acid chain to result in an immedi-
ate scission, and almost not consumed in the process.
was almost not consumed during two cycles cleavage
reactions. The same results were obtained for R4 and
R2.
2.5. Effect on proteinꢀs immunology activity
To evaluate the R5 as a DNA cleavage catalyst, we de-
signed an experiment by tracking the changes of its fluo-
rescence intensity during the catalytic reactions, which
was described in Experimental section. Agarose (1%)
gel electrophoresis showed the same DNA cleavage effi-
ciency of the compound after two cycles of DNA degra-
dation (data not shown). The spectral experiments
showed a decrease extend of the fluorescence intensity
of R5 in the process of DNA cleavage (Fig. 3a), and
the decreased rate became 6.29% (Fig. 3b) after the sec-
ond addition of DNA. However, the analysis indicated
the fluorescent quenching of R5 might mainly be due
to its binding with DNA rather than itself real consump-
tion during the photocleavage, in fact, the fluorescence
quenching was also observed when R5 was mixed with
calf thymus DNA in dark, for example, 0%, 25%,
The photoeffects of R2, R4, and R5 on a recombinant
protein fused with a 6His-Tag, was analyzed by the
method of Western blotting. The recombinant protein
was produced by Escherichia coli, which carried the gene
of a thrombin-like enzyme, gloshedobin, from the snake
venom of Gloydius shedaoensis. Western blotting ana-
lysis using anti-His-Tag serum showed positive combina-
tion of antibody and the enzyme of interest, suggesting
almost no effect on the immunology activity by using
the compounds (Fig. 4). Of course, in the control experi-
ments of the photoeffects of the compounds on amino
acids HPLC analysis suggested that the degradation of
tyrosine happened, although it was much complicated
to know how less these compounds to modify the struc-
ture of the protein.
2
3
3
5%, and 47% fluorescent quenching for R5 during addi-
2.6. Effect on protein’s enzymatic activity
tion of calf thymus DNA (0, 50, 100, 200 lM) in
DMSO–10 mM Tris–HCl (pH 7.5) solution (1:4, v/v).
Therefore, the decreasings for R5ꢀs fluorescence really
The effects of R2, R4, and R5 on enzymatic activity of
trypsin, which is a frequently used serine protease, was

Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
1619
3. Discussion
Although R2, R4, and R5 could be used as efficient tools
for the degradation of genomic DNA rather than pro-
teins, their selectivities and mechanisms were different.
R5 gave the best result: high degradation ability for
genomic DNA and no damage to proteinꢀs activity; R2
showed almost no damage to proteinꢀs activity but mod-
erate degradation ability for genomic DNA; R4 dam-
aged both to proteinꢀs activity and to genomic DNA.
These differences in the selectivity might be caused by
their different damage mechanisms. R2 and R5 photo-
degraded DNA through electron transfer mechanism
involved superoxide anion radical like co-sensitiza-
Figure 4. Western blotting analysis for the effect on recombinant
gloshedobin mixed with the compounds after phototreatment under
400 nm for 120 min. Lane 1, without compounds; lane 2, with R2, R4,
and R5, respectively.
2
0b,9
tion.
free radical mechanism involved the triplet state of tri-
For R4, the degradation was partly through
2
0c,9
diazole.
The photobioactivities of compounds were
99%
100% 100%
97%
mainly dependent on their photochemical behaviors,
the electron transfer process was usually orientation to
some targets and free radicals were disperse around
reaction site, therefore, the compounds in electron trans-
fer mechanism usually had high selectivity in photoclea-
vage by comparison with the poor selectivity of ones in
100%
93%
87%
8
2%
4%
8
7
4
0%
0%
2
0,9
free radical mechanism.
Their selectivities might be also related with their bind-
ing modes with DNA. Although there was some possi-
bility that the ring structures of R2, R5, and R4
interacted with DNA bases like intercalation, the flexi-
ble side chain of N-alkyl substituted aminoalkyl group
as acceptor at imide moieties of R5 and R2 might easily
Figure 5. Effects on specific activity of trypsin by specific precipitants
and the compounds. The enzyme (2 mg/mL) in 20 mM Tris–HCl,
pH 7.5, mixed at 25 ꢁC with 0.2 g/L specific precipitant, ammonium
sulfate (AS), cetyltrimethyl ammonium bromide (CTAB), streptomy-
cin sulfate (SS), and protamine sulfate (PS), respectively. Sixty minutes
later, the mixture centrifuged at 13,000g for 30 min. The enzymatic
activity assay performed by adding 100 lL supernatant into 2.9 mL
buffer, 20 mM Tris–HCl, pH 7.5, containing 1 mM substrate, Na-
benzoyl-L-arginine ethyl ester. The liberation of products monitored at
2
4
form hydrogen bonding with the proton at the phos-
phorous hydroxyl groups of DNA duplex–helix (Scheme
3), and R4 cannot bind with DNA in the same way. The
binding mode of hydrogen bonding for R2 and R5 made
them close to DNA more easily to mainly increase their
selectivities, as well as efficiencies partly. In addition, en-
zymes bury their catalytic amino acid residues in a bag
via self-folding hinder the approximation of nonsub-
strate molecules, therefore, this binding mode would
not affect the bioactivity of a protein.
253 nm. For compounds, the enzyme was incubated with 0.01 M
artificial compounds under 400 nm for 60 min before assay.
assayed by using Na-benzoyl-L-arginine ethyl ester as
substrate. Figure 5 shows that R2 and R5 had almost
no photodamage to the enzymatic reactivity (99% or
1
00% activity remained, respectively), while R4 showed
slight effect that caused 7% activity loss.
Although R5, R1–R3 could photocleave both plasmid
and genomic DNA via electron transfer mechanism, R3
showed the weakest bioactivity due to the possible steric
hindrance for hydrogen bonding with DNA by piperaz-
ino moiety at its side chain. R2, R5 with a proper-length
side chain to facilitate hydrogen bonding, show higher
activity. Consequently, it is not surprising that R1
photocleave DNA weakly because of its disability of
hydrogen bonding. Moreover, the reason of R4 having
stronger bioactivity, possibly be owing to its partly free
radical mechanism for photocleavage (Scheme 4).
Under the identical conditions, several frequently ap-
plied nucleic acids removal reagents involving ammo-
nium sulfate (AS), cetyltrimethyl ammonium bromide
(
CTAB), streptomycin sulfate (SS), and protamine sul-
fate (PS) were added to the enzyme solution. As shown
in Figure 5, the addition of CTAB, SS, and PS resulted
in the loss of enzymatic activities ranging from 13% to
1
8%, which was much higher than for the compounds.
Special attention is given to the precipitants, such as
AS, which can be used not only as nucleic acids precipi-
tants but also as protein precipitants. Therefore, it is not
surprising that it can be effective in removing nucleic
acids and removing proteins at the same time. In this
research, we found that only 2 g/L ammonium sulfate
will cause 3% activity loss (Fig. 5). However, in most
cases, the efficient nucleic acids removal required at least
Although we could not know how less these compounds
to modify the structure of the protein, the comparison of
impairment to an enzymatic activity by commercially
available precipitants and the synthetic compounds, sug-
gested that R2, R4, and R5 were good nucleic acids
removal reagents alternative. Under physiological con-
ditions, 2.38 · 10 M of R5 exhibited the nuclease
activity of 8Unit DNase I in the degradation of maize
genomic DNA.
ꢀ
4
70 g/L ammonium sulfate, which will cause much loss of
activity.

1620
Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
HCl (2.9 mL), and acetic acid (28mL) a solution of
sodium nitrate (8.35 g) in water (11 mL) was added
B
H
dropwise during 20 min at 0–5 ꢁC. The reaction mixture
O
was added to a solution of CuSO (8.17 g) in acetic acid
S
O
4
H
H
H
(7 mL) mixed with water (120 mL) at boiling tempera-
ture for 1.5 h to afford orange solid (3.5 g,
mp >300 ꢁC, 99% yield) of benzothioxanthene; (c)
H
O
P
O
O
B'
H
O
O
0
.3 g of benzothioxanthene was refluxed with appropri-
N
H
ate amine (0.17 mL) in ethanol (20 mL) for 2 h to afford
the product. R1: Separated on silica gel chromatography
(ethyl acetate–petroleum = 2:5, v/v), mp 185–186 ꢁC,
O
O
N
H
H
H
H
O
P
7
7
5% yield. dH (500 MHz, CDCl , Me Si) 0.98(3H, t, J
3
4
HO
O
.37, CH ), 1.43–1.50 (2H, m, (Me)–CH ), 1.68–1.78
3
2
(
2H, m, (C H )–CH ), 4.19 (2H, t, J 7.60, CONCH ),
2 5 2 2
O
7
7
7
.38–7.43 (3H, m, 8-H, 9-H, 10-H), 7.52 (1H, d, J
.98, 7-H), 8.20–8.27 (2H, m, 1-H, 6-H), 8.44 (1H, d, J
.98, 2-H), 8.64 (1H, d, J 8.13, 5-H). C H NO S re-
Scheme 4. The binding and photocleavage to DNA by compound R5.
2
2
17
2
quires: C, 73.51; H, 4.77; N, 3.90. Found C, 73.23; H,
4.69; N, 3.88. R2: Separated on silica gel chromatogra-
phy (chloroform–acetone = 1:1, v/v), 212–213 ꢁC, 85%
yield. dH (500 MHz, CDCl , Me Si) 1.20–1.34 (8H, m,
4
. Conclusion
We present novel genomic DNA photonuclease of thio-
heterocyclic naphthalimides with N-alkyl substituted
aminoalkyl side chain, which showed highly efficient
abilities in the degradation of plasmid and genomic
DNA under the mild conditions without obvious
impairment on the proteinsꢀ bioactivities. Their differ-
ences in photodegradation selectivity to DNA rather
than proteins were depended on their photodamage
mechanisms and binding modes with bio-macromole-
cules. R5 almost have no consumption after two cycles
of the photodegradation, and showed the best results
in the selectivity and efficiency of photodamage. The
structural characteristics accounting for these properties
will invoke chemists to design new types of chemical
nucleases and biologist to apply these compounds. Also,
the application potential will attract attentions of bioen-
gineers who would like to seek inexpensive and efficient
nuclease to the removal of nucleic acids from biochem-
ical and pharmaceutical preparations.
3
4
NCH , NCH ), 4.42 (2H, t, J 6.78and 6.65, CONCH ),
3
2
2
7.34–7.42 (m, 3H, 8-H, 9-H, 10-H), 7.48 (1H, d, J 7.99,
7-H), 8.14–8.20 (2H, m, 1-H, 6-H), 8.40 (1H, d, J 7.98,
2-H), 8.59 (1H, d, J 8.12, 5-H); EI-MS: m/z 374 (M ,
+
1.62), 304 (23.30), 303 (29.64), 71 (29.58), 58 (100), 56
ꢀ
1
(9.81), 43 (12.49); mmax (KBr)/cm 2950, 2870, 1695,
1660, 1560, and 1380. C H N O S requires: C, 70.57;
2
2
18
2
2
H, 4.85; N, 7.48. Found: C, 70.46; H, 4.98; N, 7.35.
R3: Separated on silica gel chromatography (chloro-
form–methanol–acetate = 9:2:1.5, v/v), 286–287 ꢁC,
81% yield. dH (500 MHz, CDCl , Me Si) 2.80 (2H, t, J
3
4
5.96 and 6.00, NCH ), 2.89 (4H, s, NHCH (cyclo)),
3.14 (4H, s, NCH (cyclo)), 4.32 (2H, t, J 6.08and
2
2
2
5.88, CONCH ), 7.38–7.46 (3H, m, 8-H, 9-H, 10-H),
2
7.54 (1H, d, J 8.06, 7-H), 8.19–8.25 (2H, m, 1-H, 6-H),
8.42 (1H, d, J 8.01, 2-H), 8.62 (1H, d, J 8.18, 5-H);
EI-MS: m/z 415(M , 5.45), 373 (39.42), 330 (35.98),
+
303 (27.95), 99 (100.0), 70 (27.95), 56 (42.09), 42
ꢀ
1
(
21.07); mmax (KBr)/cm 3410, 2970, 2840, 1690, 1640,
and 1330. C H N O S requires: C, 69.38; H, 5.09; N,
2
4
21
3
2
5
. Experimental
10.11. Found C, 69.56; H, 5.23; N, 9.96. R5: Separated
on silica gel chromatography (chloroform–etha-
nol = 5:1, v/v), 202–203 ꢁC, 78% yield. dH (500 MHz,
CDCl , Me Si) 2.46 (8H, s, NCH , NCH ), 2.70 (2H,
5
.1. Syntheses of compounds
3
4
3
2
Melting points were taken on a digital melting point
apparatus WRS-1 (Shanghai, China) and it was uncor-
rected. Infrared spectra were recorded on a Nicolet FT
t, J 7.44, CH ), 4.25 (2H, t, J 7.24, CONCH ), 7.40
2 2
(3H, m, 8-H, 9-H, 10-H), 7.49 (1H, d, J 7.98, 7-H),
8.20 (2H, m, 1-H, 6-H), 8.39 (1H, d, J 7.98, 2-H), 8.59
(1H, d, J 8.10, 5-H). C H N O S requires: C, 71.11;
1
IR-20SX, mass spectra on a Hitachi M80, H NMR
2
3
20
2
2
on a Bruker AM-300 or AM-500 using TMS as an inter-
nal standard. Combustion analysis for elemental com-
position was done on Italy MOD.1106 analyzer.
Absorption spectra were measured on Shimadzu UV-
H, 5.19; N, 7.21. Found: C, 71.23; H, 4.92; N, 7.14.
The synthesis of R4: (a) 2.32 g of 4-bromo-3-nitro-1,8-
naphthalic anhydride was reacted with benzyl mercap-
tan (0.88 mL) in the presence of K CO (0.52 g) at
2
65, fluorescence spectra on a Hitachi 850.
2
3
8
0 ꢁC for 8h under Ar gas protection. The reaction mix-
The syntheses of R1–R3, and R5: (a) 4-Bromo-1,8-naph-
thalic anhydride (4.155 g), o-aminobenzenethiol (2.12 g),
and K CO (1.035 g) were mixed and refluxed in DMF
ture was poured into salted-ice water to afford yellow
solid (2.47 g, mp 188–193 ꢁC, 94% yield) of 4-
(benzylthioxy)-3-nitro-1,8-naphthalic anhydride; (b) this
intermediate was reduced by SnCl Æ2H O (7.6 g) in HCl
2
3
(
42 mL) for 30 min, then the mixture was poured into
2
2
water (100 mL), acidified, and filtered to afford green
solid (3.71 g, mp 198–203 ꢁC, 77% yield) of 4-(2-ami-
no-phenylthioxy)-1,8-naphthalic anhydride; (b) to the
mixture of this intermediate (3.71 g), water (4.6 mL),
(31 mL) at 90 ꢁC for 2 h to give yellow solid (2.85 g, mp
171–185 ꢁC); (c) to the mixture of the amino intermedi-
ate (1.46 g), acetic acid (57 mL), water (7 mL), and HCl
(8.5 mL), a solution of NaNO (0.31 g) in water (10 mL)
2

Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
1621
was added dropwise during 20 min at 0 ꢁC and the reac-
5.3. Nuclease activity assay
tion mixture was stirred for 30 min at 0 ꢁC and for 7 h at
1
8
5–20 ꢁC, yellowish green solid (0.97 g, mp 162–167 ꢁC,
6% yield) of 5-oxa-10-thia-8,9-diaza-cyclopenta[a]-
Photosensitive compound (5 lL, 10 mM) dissolved in
DMSO was added into a reaction mixture containing
500 ng freshly prepared maize genomic DNA in 15 lL
20 mM Tris–HCl, pH 7.5. The reaction was initiated
with light activation under 400 nm and stopped by the
removal of the light at different time points. Nuclease
activity was determined by monitoring the increase
in the absorbance at 260 nm. When 3 lL of bovine
phenalene-4,6-dione was obtained; (d) then this interme-
diate (0.38g) was reacted with butylamine (0.30 mL) in
absolute ethanol (20 mL) at reflux for 2 h to afford yel-
low solid in 72% yield of R4, which was purified through
silica gel chromatography (chloroform–acetate = 200:1,
v/v). R4: mp 177–179 ꢁC.
1
H NMR (500 MHz,
DMSO-d , Me Si): d (ppm) 0.94 (t, J1 7.36, 3H,
pancreas DNase
3.1 · 10 Da) (Takara, Japan) was applied as standard,
the reaction buffer was changed to 210 lL 20 mM
I
(70 U/lL, 20.4 mg/mL, Mw
6
4
H
4
CH3), 1.38(m, 2H, (CH 3)CH2), 1.64 (m, 2H,
(
C H )CH ), 4.05 (t, J 7.49, 2H, CONCH ), 8.03 (t, J
2 5 2 2
7
7
.77, 1H, 2-H), 8.62 (d, J 7.11, 1H, 1-H), 8.82 (t, J
.97, 1H, 3-H), 9.32 (s, 1H, 7-H). HR-MS:
Tris–HCl, pH 7.5, containing 4 mM MgCl and 1 mM
2
dithiothreitol. One unit was defined as the amount of
nuclease that causes an increase in absorbance at
260 nm of 0.001 per minute per milliliter at 25 ꢁC,
pH 7.5, with highly polymerized DNA as the
C H N O S require: 311.0728. Found: 311.0728. EI-
2
1
6
13
3
+
MS: m/z (%): 311.0728(M , 17.00), 283.0705 (100.00),
2
1
1
1
66.0659 (27.86), 241.0200 (28.75), 227.0069 (89.84),
1
8
57.0128(23.66). IR (KBr): 2950, 2 78 0, 1710, 1660,
300 cm . C H N O S require: C 61.72, H 4.21, N
3.50. Found: C, 61.59; H, 4.41; N, 13.62.
substrate.
ꢀ
1
1
6
13
3
2
5.4. Fluorescence detection assay
5.2. Total DNA isolation and purification from the leaf of
maize
Photosensitive compound (5 lL, 10 mM) dissolved in
DMSO was added into a reaction mixture containing
500 ng freshly prepared maize genomic DNA in 15 lL
The extraction and purification of total DNA was per-
formed by the method of CTAB. The CTAB isolation
buffer (4 M NaCl, 100 mM Tris–HCl, pH 8.0, 20 mM
EDTA, 20 g/L ethyltrimethylammonium bromide,
20 mM Tris–HCl, pH 7.5. The reaction was initiated
with light activation under 400 nm and stopped by the
removal of the light after 20 min. The reaction mixture
was diluted until the concentration of the compound
at 1 · 10 M for fluorescence detection assay. The
emission wavelength is at 470 nm and the absorption
at 522 nm. For consumption assay of the compound
during the cleavage reaction, this procedure was re-
peated with another addition of 500 ng DNA into the
reaction mixture. The other operation conditions re-
mained the same.
ꢀ5
2
mL/L 2-mercaptoethanol) was preheated at 60 ꢁC.
Fresh leaf tissue (2 g) of maize was ground to a powder
in liquid nitrogen in a chilled mortar and pestle. Then
the powder was scraped into a chilled 50 mL tube.
CTAB (5 mL) buffer per gram tissue was added into
the tube and then incubated the sample at 60 ꢁC for
3
0 min with occasional gentle swirling. The same volume
of phenol–chloroform (1:1, v/v) was added to the sample
with gently swirling. The precipitates were discarded
after centrifugation at 16,000g at 25 ꢁC for 10 min, while
the yellow, aqueous phase with a wide-bore pipette was
transferred to a new 50 mL tube. Two-thirds volume of
cold 100% isopropanol was added to the tube and mixed
gently to precipitate the nucleic acids. The precipitate
was collected and then incubated for a minimum of
5.5. Enzymatic activity assay
Enzymatic activity of trypsin (Ameresco, USA) was as-
sayed using the substrate, Na-benzoyl-L-arginine ethyl
ester (BAEE) (Sigma). In brief, trypsin (2 mg/mL,
500 U) in 20 mM Tris–HCl, pH 7.5, was mixed at
25 ꢁC with 0.2 g/L specific precipitant, ammonium sul-
fate (AS), cetyltrimethyl ammonium bromide (CTAB),
streptomycin sulfate (SS), and protamine sulfate (PS),
respectively. Sixty minutes later, the mixture was centri-
fuged at 13,000g for 30 min. The enzymatic activity
assay was performed by adding 100 lL of supernatant
into 2.9 mL of buffer, 20 mM Tris–HCl, pH 7.5, con-
taining 1 mM BAEE. The liberation of products was
monitored at 253 nm, and the enzyme was incubated
with 0.01 M artificial compounds under 400 nm for
60 min before assay.
2
0 min with 75% ethanol. The DNA precipitate was ob-
tained again by centrifugation at 16,000g for 10 min and
then was left to dry in the air. The DNA pellet was
resuspended in 2–3 mL TE buffer (1 M Tris–HCl,
0
.5 M EDTA, pH 8.0) and then RNase A was added
to the mixture at a final concentration of 10 lg/mL to re-
move RNA for 30 min at 37 ꢁC. Finally, the RNase A
was removed by adding an equal volume of phenol–
chloroform (1:1, v/v) into the mixture followed with cen-
trifugation at 16,000g at 25 ꢁC for 10 min. The upper,
aqueous phase was collected and then 7.5 M ammonium
acetate, pH 7.7 to a final concentration of 2.5 M and 2.5
volume of ice-cold 100% ethanol was added into the
supernatant to precipitate the DNA. Centrifuged at
5.6. Western blotting assay
The total cellular proteins after being treated with the
compounds were analyzed by 12.5% SDS-PAGE. The
protein bands were electrophoretically transferred onto
nitrocellulose membrane at 0.65 mA/cm for 2 h. The
membrane was blocked with 20 g/L fat-free milk and
incubated with horse anti-His-Tag serum (Invitrogen,
1
0,000g for 10 min at 4 ꢁC, the DNA pellet was obtained
and then was washed with 75% ethanol followed with
centrifugation at 10,000g for 10 min at 4 ꢁC. The
DNA pellet was resuspended in 50 lL of TE buffer
2
(
1 M Tris–HCl, 0.5 M EDTA, pH 8.0) for future usage.

1622
Q. Yang et al. / Bioorg. Med. Chem. 13 (2005) 1615–1622
USA) diluted 1:300, followed by rabbit anti-horse IgG-
HRP conjugate (Invitrogen, USA) diluted 1:500. The
bound antibody was detected by using 3,3 -diamino-
benzidine and hydrogen peroxide.
10. Franklin, S. J. Curr. Opin. Chem. Biol. 2001, 5, 201–
08.
11. Cowan, J. A. Curr. Opin. Chem. Biol. 2001, 5, 634–
42.
2. Wan, R.; Zhao, G.; Chen, J.; Zhao, Y. Chin. Sci. Bull.
000, 45, 2017–2028.
3. Wilcox, D. E. Chem. Rev. 1996, 96, 2435–2458.
2
0
6
1
1
2
Acknowledgements
14. Sigman, D. S.; Mazumder, A.; Perrin, D. M. Chem. Rev.
993, 93, 2295–2316.
1
1
5. Branum, M. E.; Tipton, A. K.; Zhu, S.; Que, L. J. J. Am.
Chem. Soc. 2001, 123, 1898–1904.
6. Sissi, C.; Rossi, P.; Felluga, F.; Formaggio, F.; Palumbo,
M.; Tecilla, P.; Toniolo, C.; Scrimin, P. J. Am. Chem. Soc.
This study was under the auspices of the National Key
Project for Basic Research and 863 Project for High-
Tech, the Ministry of Science and Technology of China
1
(
2003CB114405 and 2003AA2Z3520).
2
001, 123, 3169–3170.
1
1
7. Neves, A.; Terenzi, H.; Horner, R.; Horn, A. J.; Szpo-
ganicz, B.; Sugai, J. Inorg. Chem. Commun. 2001, 4, 388–
391.
8. Kumar, C. V.; Buranaprapuk, A.; Opiteck, G. J.; Moyer,
M. B.; Jockusch, S.; Turro, N. J. Proc. Natl. Acad. Sci.
U.S.A. 1998, 95, 10361–10366.
References and notes
1
. Klein, J.; Altenbuchner, J.; Mattes, R. J. Biotechnol. 1998,
0, 145–153.
. Zylicz-Stachula, A.; Harasimowicz-Słowi n´ ska, R. I.;
Sobolewski, I.; Skowron, P. M. Nucleic Acids Res. 2002,
6
_
2
19. Kumar, C. V.; Buranaprapuk, A.; Thota, J. Proc. Indian
Acad. Sci. (Chem. Sci.) 2002, 114, 579–592.
3
0, 733.
20. (a) Qian, X.; Huang, T., et al. J. Chem. Soc., Perkin
Trans. 2 2000, 715–718; (b) Qian, X.; Li, Y.; Xu, Y.; Liu,
Y.; Qu, B. Bioorg. Med. Chem. Lett. 2004, 14, 2665–2668;
(c) Li, Y.; Xu, Y.; Qian, X.; Qu, B. Bioorg. Med. Chem.
Lett. 2003, 13(20), 3513–3515.
3
4
5
. Jiang, C. Z.; Yee, J.; Mitchell, D. L.; Britt, A. B. Proc
Natl. Acad. Sci. U.S.A. 1997, 94, 7441–7445.
. Bellamy, S. R. W.; Baldwin, G. S. Nucleic Acids Res. 2001,
29, 3857–3863.
. Goulas, E.; Dily, F. L.; Simon, J. C.; Ourry, A. J. Exp.
Botany 2002, 53, 1941–1948.
21. Brana, M.; Ramos, A. Curr. Med. Chem.: Anti-cancer
Agents 2001, 1, 237–255.
6
7
. Kramer, R. Coord. Chem. Rev. 1999, 182, 243–261.
. Franke, I.; Meiss, G.; Blecher, D.; Gimadutdinow, O.;
Urbanke, C.; Pingoud, A. FEBS Lett. 1998, 425, 517–522.
. Rangarajan, S.; Shankar, V. Biochim. Biophys. Acta 1999,
22. Kunitz, M. J. Gen. Physiol. 1950, 33, 349–355.
23. Yang, Q.; Xu, J. Q.; Li, M.; Lei, X. Y.; An, L. J.
Biotechnol. Lett. 2003, 25, 607–610.
24. DeSilva, A. P.; Gunaratne, H. Q. N.; Habib-Jiwan, J.-L.;
McCoy, C. P.; Rice, T. E.; Soumiliion, J.-P. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1728.
8
9
1473, 293–304.
. Armitage, B. Chem. Rev. 1998, 98, 1171–1200.