Synthesis of methyl-substituted xanthotoxol to clarify prooxidant effect of methyl on radical-induced oxidation of DNA

Article abstract of DOI:10.1016/j.ejmech.2010.02.044

4-Methyl-8-hydroxylpsoralen (MXan) and 4,9-dimethyl-8-hydroxylpsoralen (DMXan) were synthesized in order to clarify the effect of methyl on the antioxidant effectiveness of xanthotoxol (8-hydroxylpsoralen, Xan), which were assessed by bleaching β-carotene

Full text of DOI:10.1016/j.ejmech.2010.02.044

European Journal of Medicinal Chemistry 45 (2010) 2559e2566
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis of methyl-substituted xanthotoxol to clarify prooxidant effect of
methyl on radical-induced oxidation of DNA
*
Chuan Xiao, Zhi-Guang Song, Zai-Qun Liu
Department of Organic Chemistry, College of Chemistry, Jilin University, No. 2519 Jiefang Road, Changchun 130021, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
4-Methyl-8-hydroxylpsoralen (MXan) and 4,9-dimethyl-8-hydroxylpsoralen (DMXan) were synthesized
in order to clarify the effect of methyl on the antioxidant effectiveness of xanthotoxol (8-hydroxyl-
Received 2 September 2009
Received in revised form
8 February 2010
Accepted 18 February 2010
Available online 4 March 2010
psoralen, Xan), which were assessed by bleaching
b-carotene in linoleic acideTriton emulsion, by
interacting with 2,20-azinobis(3-ethylbenzothiazoline-6-sulfonate) cationic radical (ABTSþꢀ), 2,20-
diphenyl-1-picrylhydrazyl radical (DPPH), and galvinoxyl radical, and by protecting DNA against the
oxidation induced by Cu2þ/glutathione (GSH) and 2,20-azobis(2-amidinopropane hydrochloride) (AAPH).
Methyl attaching to xanthotoxol did not affect its ability to protect linoleic acid against autoxidation and
to inhibit Cu2þ/GSH-induced oxidation DNA, but decreased its ability to scavenge ABTSþꢀ and DPPH, and
to protect DNA against AAPH-induced oxidation. Therefore, methyl attenuated the antioxidant effec-
tiveness of xanthotoxol in radical-induced oxidation of DNA.
Keywords:
Xanthotoxol
Free radical
Antioxidant
Ó 2010 Elsevier Masson SAS. All rights reserved.
Oxidation damage of DNA
1. Introduction
was applied to condense with anhydride for preparing antitumor
drugs [12]. As an alkyl group, the effect of methyl on the bioactivity
Xanthotoxol (8-hydroxylpsoralen, Xan) was a natural psoralen
[1] with high ability to treat Alzheimer disease [2], to inhibit human
cytochrome P450 [3], and to retard the formation of fibril and
of psoralens was not often reported. We have reported that methyl
attaching to 4-position in coumarin played a prooxidant role in the
oxidation of low-density lipoprotein (LDL) induced by 2,20-azobis
b
-amyloid [4]. Because the hydroxyl at 8-position made xanthotoxol
(2-amidinopropane
hydrochloride)
(AAPH,
ReN]NeR,
a free-radical-scavenger obviously [5,6], a large amount of research
dealt with the isolation of xanthotoxol derivatives from plants [7]
and the investigation of antioxidant effectiveness. However,
8-methoxypsoralen increased the amount of 8-hydroxy-20-deoxy-
guanosine (8-OHdG) when calf thymus DNA was irradiated by
ultraviolet A (UVA) [8]. The cytotoxicity of 8-methoxypsoralen was
due to induce the formation of interstrand cross-link of DNA under
irradiation of UVA [9], in which the conversion among various
excited states of 8-methoxypsoralen was confirmed by quantum
calculation [10].
R ¼ eCMe2C(]NH)NH2) [13]. So, presented here was a study to test
whether methyl also played prooxidant role in xanthotoxol. As
shown in Scheme 1, 4-methyl-8-hydroxylpsoralen (MXan) and 4,9-
dimethyl-8-hydroxylpsoralen (DMXan) were chosen to explore the
effect of methyl at furan ring and/or lactone ring on the antioxidant
effectiveness of xanthotoxol, the routines to synthesize MXan and
DMXan were outlined in Scheme 2 [14].
3. Pharmacology
Because a large body of evidence has revealed the correlation of
the in vivo oxidative stress with ageing [15] and some fatal diseases
[16,17], many works focused on the synthesis of antioxidants and
the evaluation of their activities in various experimental systems
[18]. Cu2þ/glutathione (GSH)-mediated and AAPH-induced oxida-
tion of DNA were usually employed as the experimental systems to
investigate the antioxidant capacity because Cu2þ may cause the
damage of DNA in the presence of similar concentration of GSH in
vivo [19], and peroxyl radical (ROOꢀ) generated from the decom-
position of AAPH can mimic the in vivo oxidative stress of DNA [20].
The effects of DMXan, MXan and Xan were evaluated on Cu2þ/GSH-
and AAPH-induced oxidation of DNA. In addition, their antioxidant
2. Chemistry
In addition to many contributions to the total synthesis of
psoralen derivatives, some works devoted to investigate the effect
of substituents on the bioactivity of psoralen. For example,
5-hydroxylpsoralen was synthesized to detect the ability of
hydroxyl to trap radicals [11], and 5-amino-8-methoxypsoralen
* Corresponding author. Tel.: þ86 431 88499174; fax: þ86 431 88499159.
E-mail address: zaiqun-liu@jlu.edu.cn (Z.-Q. Liu).
0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2010.02.044

2560
C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
4. Results and discussion
4.1. Synthesis of methyl-substituted xanthotoxol
Methyl at 4-position in methyl-substituted xanthotoxol was
derived from compound 1 employed as the reagent herein. So, the
synthesis protocol mainly focused on the construction of furan ring
in this work. The furan ring was formed by the condensation
between C]O in formacylmethoxy or acetylmethoxy and H atom at
6-position in coumarin under basic condition. DMXan was derived
from the coumarin with acetylmethoxy at 7-position, while MXan
was derived from the coumarin with formacylmethoxy at the same
position. The coumarin containing acetylmethoxy at 7-position, 10,
was prepared by using CH3COCH2Br to react with the hydroxyl in
compound 9, and the reaction between (CH3O)2CHCH2Br and
compound 1 formed the coumarin containing formacylmethoxy at
the same position, 2. However, the difficulty in the synthesis of
MXan and DMXan was to preserve the hydroxyl at 8-position when
the anion of hydroxyl at 7-position reacted with CH3COCH2Br or
(CH3O)2CHCH2Br. Williamson reaction can take place between
hydroxyl at both 7- and 8-position with CH3COCH2Br or
(CH3O)2CHCH2Br without selectivity, thus, protecting hydroxyl at
8-position to make hydroxyl at 7-position react with CH3COCH2Br
or (CH3O)2CHCH2Br was the key strategy in the synthesis of
methyl-substituted xanthotoxol. As shown in Scheme 2, compound
Scheme 1. Structures of psoralen derivatives employed in this work.
capacities were screened by bleaching b-carotene in the emulsion
of linoleic acid (LH) and Triton [21], and by interacting with
2,20-azinobis(3-ethylbenzothiazoline-6-sulfonate) cationic radical
(ABTSþꢀ) [11], 2,20-diphenyl-1-picrylhydrazyl (DPPH) [11], and
galvinoxyl radical [22], respectively.
Scheme 2. Synthesis routines of MXan and DMXan.

C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
2561
1 was chosen to be reagent because acetyl can be converted into
hydroxyl group via BaeyereVilliger oxidation and hydrolysis reac-
tion [23]. Furthermore, we have attempted to simplify the synthetic
routines as shown in Scheme 3. After 7-formacylmethoxy and 7-
acetylmethoxycoumarin, 12 and 13, were prepared by the reaction
between compound 1 and (CH3O)2CHCH2Br or CH3COCH2Br, they
were treated by NaOH in order to construct furan ring directly. But
this operation did not generate target product, 14 and 15.
antioxidant capacity in this case. On the other hand, as shown in
panel B of Fig. 1, the addition of 40.0
mM Xan decreased the
percentage of ABTSþ rapidly, while the same concentration of
DMXan and MXan cannot decrease the percentage of ABTSþꢀ as fast
as Xan, indicating that Xan possessed stronger ability to reduce
ABTSþꢀ. The decay of the percentage for ABTSþꢀ in the presence of
ꢀ
DMXan was even slower than MXan, revealing that methyl at 4-
and 9-position in xanthotoxol was not beneficial to reduce ABTSþ
.
ꢀ
As shown in Scheme 2, compound
1
reacted with
As shown in panel C of Fig. 1, the additions of 200 mM DMXan,
(CH3O)2CHCH2Br to form compound 2, and then, the acetyl con-
verted into hydroxyl group to form compound 3 by BaeyereVilliger
oxidation and hydrolysis reaction. After the hydroxyl in compound
MXan, Xan decreased the percentage of DPPH markedly. In
particular, MXan and Xan decreased the percentage of DPPH even
faster than DMXan, indicating that methyl at 4- and 9-position in
xanthotoxol was not beneficial for hydrogen atom in hydroxyl to
donate to N-centered radical. As shown in panel D of Fig. 1, both
3
was methylated to form compound 4, and ketal in
(CH3O)2CHCH2Oe was hydrolyzed to form formacylmethoxy. The
cyclization reaction occurring in compound 5 formed furan ring in
compound 6. On the other hand, the aforementioned routine was
not suitable to synthesize DMXan because BaeyereVilliger oxida-
tion will take place at 8-acetyl and 7-acetylmethoxy simulta-
neously as shown in Scheme 4. Thus, hydroxyl at 7-position in
compound 1 was protected by MOMCl, and then, 8-acetyl was
converted into hydroxyl to generate compound 8. After the
hydroxyl at 8-position in compound 8 was etherized by PhCH2Br,
the protective group of 7-OH (methoxymethoxy) was hydrolyzed to
form hydroxyl under acidic condition. Thus, the furan ring in
DMXan was constructed on the basis of the hydroxyl at 7-position
in compound 9. Finally, benzyloxy at 8-position in compound 11
was removed by hydrogenation.
DMXan, MXan and Xan (200
mM) were able to decrease the
percentage of galvinoxyl radical. The line of DMXan was almost
superposition to that of MXan in this case, revealing that DMXan
and MXan possessed the similar ability to donate hydrogen atom in
hydroxyl to O-centered radical. But Xan decreased the percentage
of galvinoxyl radical even slower than DMXan and MXan, indi-
cating that methyl was beneficial for hydrogen atom in hydroxyl to
donate to O-centered radical. Moreover, the concentrations of
DMXan, MXan and Xan to scavenge 50% ABTSþꢀ, DPPH and galvi-
noxyl radical (IC50) were measured and listed in Table 1.
A low value of IC50 implicated that the antioxidant possessed
relative high effectiveness to scavenge the corresponding radical
[24]. It can be found that DMXan, MXan and Xan were effective
scavengers to trap ABTSþ and DPPH. The lowest value of IC50 was
ꢀ
found in Xan rather than methyl-substituted xanthotoxol,
implying that introducing methyl to xanthotoxol cannot enhance
4.2. Abilities of DMXan, MXan and Xan to protect linoleic acid and
to scavenge radicals
the free-radical-scavenging ability to trap ABTSþ and DPPH.
ꢀ
Finally, high value of IC50 for Xan to trap galvinoxyl radical
revealed a completely reverse result, which was worthy to study
in the future work.
As shown in Fig. 1, the antioxidant abilities of DMXan, MXan and
Xan were screened by bleaching
scavenging ABTSþꢀ, DPPH, and galvinoxyl radicals.
LH and -carotene can form a water-soluble emulsion in the
b-carotene in LH emulsion, and by
b
presence of Triton X-100 or Tween [21]. Oxygen in atmosphere
oxidized LH to form peroxyl radical (LOOꢀ), which can be trapped by
4.3. Abilities of DMXan, MXan and Xan to protect DNA against the
oxidation induced by Cu2þ/GSH or AAPH
b
-carotene (lmax ¼ 460 nm). So, as shown as the line from the
control experiment in panel A of Fig. 1, the decay of the absorbance
at 460 nm indicated that more LOOꢀ generated from the autoxi-
The oxidation of Cu(II)-binding DNA catalyzed by intracellular
GSH increased the amount of 8-oxo-7,8-dihydro-20-deoxy-
guanosine (8-oxodG) [25] because of the generation of GSH radical
(GSꢀ) via Cu(II) þ GSH / Cu(I) þ GSꢀ [26]. The further oxidation of
DNA mediated by Cu2þ/GSH broke the strand of DNA to form pro-
penals, which can be detected after treated by thiobarbituric acid
(TBA), thus, propenals were also called TBA reactive substance
(TBARS, lmax ¼ 535 nm) [19]. In addition, AAPH usually acted as
peroxyl radical (ROOꢀ) resource to mimic LDL [27], erythrocytes
[28], and DNA [20] under oxidative stress in vitro. The oxidation of
dation of LH were eliminated by
period increasing. As shown in panel A, the decay of the absorbance
was retarded by the addition of 13.3 M DMXan, MXan or Xan,
indicating that DMXan, MXan and Xan were able to protect LH
against the autoxidation. The lines in the presence of DMXan and
MXan were almost superposition to that of Xan, revealing that the
abilities of DMXan and MXan to protect LH were the same as that of
Xan. So, methyl attaching to xanthotoxol did not affect the
b-carotene with the incubation
m
Scheme 3. Furan rings in MXan and DMXan cannot be constructed from 12 and 13.

2562
C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
Scheme 4. Possible positions for BaeyereVilliger oxidation occurring in 8-acetyl-7-acetylmethoxycoumarin.
aforementioned biological samples can also be detected by
measuring TBARS [8]. Hence, we herein applied above in vitro
experimental systems to assess the effect of DMXan, MXan and Xan
on DNA. Fig. 2 illustrated the variety of the absorbance for TBARS in
the process of the oxidation of DNA.
The percentages of oxidative DNA in the presence of DMXan,
MXan and Xan were around 85% without significant difference in
Cu2þ/GSH-induced oxidation of DNA, implying that methyl did not
affect the protective effects of xanthotoxol obviously. On the other
hand, in AAPH-induced oxidation of DNA, the addition of Xan and
MXan decreased the oxidative percentage of DNA to 60% and 68%,
respectively, while DMXan can only decrease the oxidative
percentage of DNA to 80%, indicating that methyl attenuated the
antioxidant capacity of xanthotoxol in this case. This finding was in
agreement with our previous report on the prooxidant effect of 4-
methylcoumarin in AAPH-induced oxidation of LDL [13], in which
the prooxidant effect of methyl at 4-position of coumarin was
illustrated by Scheme 5.
The hydrogen atom of methyl in 4-methylcoumarin was
abstracted by ROOꢀ to form a benzyl radical, which was able to
initiate the oxidation of LDL [13]. Similarly, hydrogen atom in methyl
at 4- and 9-position in xanthotoxol was possibly abstracted by ROOꢀ
to form corresponding radical, which was able to improve the
oxidation of DNA. But the hydroxyl at 8-position in xanthotoxol
functioned as an antioxidative group to protect DNA against AAPH-
induced oxidation. So, the observed property of methyl-substituted
xanthotoxol was the overall effect from the antioxidant property of
hydroxyl and the prooxidant property of methyl.
The line from blank experiment in Fig. 2 exhibited an increase of
the absorbance of TBARS, indicating that more propenals were
generated in the presence of Cu2þ/GSH or AAPH with the incuba-
tion period increasing. Comparing with the line from the control
experiment, the addition of DMXan, MXan and Xan moved down
the lines of the formation of TBARS, revealing that DMXan, MXan
and Xan were able to protect DNA against Cu2þ/GSH-induced
damage. Moreover, in order to compare the ability of DMXan, MXan
and Xan to protect DNA, the absorbance at 150 min in the control
experiment of Cu2þ/GSH-induced oxidation of DNA was designated
as A0, and the absorbance at the same time point in the presence of
DMXan, MXan and Xan was designated as Ax. Thus, the oxidative
percentage of DNA in the presence of DMXan, MXan and Xan can be
calculated by (Ax/A0) ꢁ 100. A low percentage of the oxidative DNA
indicated that the antioxidant possessed high ability to protect DNA
against the oxidation. Meanwhile, 360 min was chosen as the time
point to calculate the oxidative percentage in AAPH-induced
oxidation of DNA. The results were illustrated in Fig. 3.
M DMXan, MXan and Xan (A), percentage of ABTSþ scavenged by 40.0
mM DMXan, MXan and
ꢀ
Fig. 1. Decay of the absorbance of
b-carotene-LH emulsion in the presence of 13.3
m
Xan (B), percentage of DPPH scavenged by 200
mM DMXan, MXan and Xan (C), and percentage of galvinoxyl radical scavenged by 200
mM DMXan, MXan and Xan (D).

C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
2563
Table 1
IC50 of DMXan, MXan and Xan in scavenging ABTSþ , DPPH and galvinoxyl radical.
which BrCH2CH(OCH3)2 (14.8 mL, 98.4 mmol) was added drop-
wisely at room temperature. Then, the mixture was heated to
140 ꢂC and stirred for 4 h. After cooled to room temperature, the
mixture was diluted by water, and then, extracted by ethyl acetate.
The organic phases were washed by brine, dried over Na2SO4. After
the organic solvent was evaporated, the crude product was purified
by flash chromatograph column with petroleum ether and ethyl
acetate (1:1, v:v) as eluent. Compound 2 was obtained as a white
solid, yield 86%. 1H NMR (300 MHz, CDCl3): 7.56 (d, 1H, J ¼ 9.0 Hz, H
at 5-position), 6.89 (d, 1H, J ¼ 9.0 Hz, H at 6-position), 6.17 (s, 1H, H
at 3-position), 4.68 (s, 1H, (CH3O)2CHCH2Oe), 4.10 (d, 2H, J ¼ 5.1 Hz,
(CH3O)2CHCH2Oe), 3.46 (s, 6H, (CH3O)2CHCH2Oe), 2.61 (s, 3H,
CH3eC]O), 2.40 (s, 3H, eCH3 at 4-position).
ꢀ
Antioxidants
IC50 (mM)
ABTSþ
DPPH
Galvinoxyl
ꢀ
DMXan
MXan
Xan
18.0
15.8
10.3
118.8
81.5
69.6
170.1
173.2
188.3
5. Conclusion
Methyl attaching to xanthotoxol did not influence the protective
effect on the autoxidation of LH and on the oxidation of DNA
induced by Cu2þ/GSH, but decreased the ability to scavenge ABTSþ
ꢀ
and DPPH radicals, and to protect DNA against AAPH-induced
oxidation. So, methyl attenuated the antioxidant effectiveness of
hydroxyl in xanthotoxol, and the prooxidant mechanism of methyl
in xanthotoxol will be explored experimentally in the further
research. The obtained results give a useful caution for the desig-
nation of antioxidant because benzyl-type methyl has the potential
to initiate additional radical propagation.
6.2.2. 4-Methyl-7-(20,20-dimethoxyethoxy)-8-hydroxylcoumarin (3)
Compound 2 (21.4 g, 70 mmol) was heated in 88 mL of 2 N NaOH
aqueous solution for 45 min, then cooled to 0 ꢂC, to which hydrogen
peroxide (90 mL, 30 wt%) was added dropwisely under stirring. The
mixture was stirred for 2 h, and acidified by 250 mL of 2 N HCl
aqueous solution to pH ¼ 1 at 0 ꢂC. Compound 3 was precipitated
and yield 72%. 1H NMR (300 MHz, CDCl3): 7.08 (d, 1H, J ¼ 8.7 Hz, H
at 5-position), 6.90 (d, 1H, J ¼ 8.7 Hz, H at 6-position), 6.18 (s, 1H, H
at 3-position), 4.78 (s, 1H, (CH3O)2CHCH2Oe), 4.14 (d, 2H, J ¼ 5.1 Hz,
(CH3O)2CHCH2Oe), 3.48 (s, 6H, (CH3O)2CHCH2Oe), 2.39 (s, 3 H,
eCH3 at 4-position).
6. Experimental protocols
6.1. Materials
AAPH, diammonium salt of 2,20-azinobis(3-ethylbenzothiazoline-
6-sulfonate) (ABTS), DPPH, naked DNA sodium salt, GSH, LH, b-caro-
6.2.3. 4-Methyl-7-(20,20-dimethoxyethoxy)-8-methoxycoumarin (4)
Methyl iodide (0.75 mL, 12 mmol) and K2CO3 (2.76 mg,
20 mmol) was dissolved in 150 mL of DMF, to which compound 3
(2.8 g, 10 mmol) was added. The mixture was stirred for 6 h at room
temperature, then CHCl3 and water were added. The aqueous phase
was extracted by CHCl3. The organic phase was combined and
washed by water for three times, and dried over Na2SO4. After the
organic solvent was evaporated, the crude product was purified by
flash chromatograph column with petroleum ether and ethyl
acetate (3:1, v:v) as eluent. Compound 4 was obtained as a white
solid powder, yield 95%. 1H NMR (300 MHz, CDCl3): 7.27 (d, 1H,
J ¼ 7.5 Hz, H at 5-position), 6.89 (d, 1H, J ¼ 7.5 Hz, H at 6-position),
6.16 (s, 1H, H at 3-position), 4.77 (s, 1H, (CH3O)2CHCH2Oe), 4.12 (d,
2H, J ¼ 5.1 Hz, (CH3O)2CHCH2Oe), 3.98 (s, 3H, eOCH3 at 8-position),
3.48 (s, 6H, (CH3O)2CHCH2Oe), 2.39 (s, 3H, eCH3 at 4-position).
tene, and 8-methoxypsoralen were purchased from ACROS
ORGANICS, Belgium, and used as received. Xanthotoxol was derived
from the demethylation of 8-methoxypsoralen [29], and 13C NMR
(DMSO-d6, 75 MHz) of Xan were found at 159.6, 147.0, 145.3, 145.0,
139.7, 129.8, 124.9, 116.0, 113.5, 109.9, 106.7. 4-Methyl-7-hydroxyl-8-
acetylcoumarin (1) was synthesized following literature [30]. Other
reagents were of analytical grade, and purchased from Beijing
Chemical Reagent Co., China. Flash column chromatography was
performed on a silica gel (300e400 mesh) and thin layer chroma-
tography (TLC) inspections were carried out on a silica gel GF254
plates. The structures of the obtained compounds were identified by
1H NMR (Varian Mercury 300 NMR spectrometer). The synthetic
compounds were analyzed by high performance liquid chromatog-
raphy, and the purities of these compounds were larger than 98.0%.
6.2.4. 4-Methyl-7-formylmethyl-8-methoxycoumarin (5)
6.2. Synthesis of 4-methyl-8-hydroxylpsoralen (MXan)
Compound 4 (0.88 g, 3 mmol) was refluxed in 4 mL of 5% HCl
for 10 min under nitrogen atmosphere. After the mixture was
cooled, the precipitate was purified by flash chromatography
column with petroleum ether and ethyl acetate (1:2, v:v) as eluent.
6.2.1. 4-Methyl-7-(20,20-dimethoxyethoxy)-8-acetylcoumarin (2)
Compound 1 (19.2 g, 88 mmol) and K2CO3 (22.6 g, 164 mmol)
was dissolved in 100 mL of N,N-dimethylformamide (DMF), to
B
A
Fig. 2. The variety of the absorbance of TBARS in the oxidation of DNA (2.0 mg/mL) induced by 5.0 mM Cu2þ and 3.0 mM GSH in the presence of 0.2 mM DMXan, MXan or Xan (A),
and by 40 mM AAPH and in the presence of 80.0 M DMXan, MXan or Xan (B).
m

2564
C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
6.3. Synthesis of 4,9-dimethyl-8-hydroxylpsoralen (DMXan)
Cu(II) and GSH
AAPH
100
85
6.3.1. 4-Methyl-7-methoxymethoxy-8-acetylcoumarin (7)
Compound 1 (19.2 g, 88 mmol), MOMCl (12.0 mL, 132 mmol)
and K2CO3 (24.3 g,176 mmol) were dissolved in 200 mL of DMF, and
stirred at 40 ꢂC till compound 1 cannot be detected by TLC. Then,
the mixture was poured into ice-water with stirring. Compound 7
was precipitated and yield 85%. 1H NMR (300 MHz, CDCl3): 7.55 (d,
J ¼ 8.7 Hz, 1H, H at 5-position), 7.12 (d, J ¼ 8.7 Hz, 1H, H at 6-
position), 6.18 (s, 1H, H at 3-position), 5.26 (s, 2H, CH3OCH2Oe),
3.49 (s, 3H, CH3OCH2Oe), 2.62 (s, 3H, CH3C]O), 2.41 (s, 3H, eCH3 at
4-position).
70
6.3.2. 4-Methyl-7-methoxymethoxy-8-hydroxycoumarin (8)
Compound 7 (18.3 g, 70 mmol) was stirred in 88 mL of 2 N NaOH
aqueous solution at 0 ꢂC, to which hydrogen peroxide (90 mL, 30 wt
%) was added dropwisely. The mixture was stirred for 2 h, and then,
acidified by 250 mL of 2 N HCl aqueous solution to pH ¼ 1 at 0 ꢂC.
Compound 8 was precipitated and yield 73%. 1H NMR (300 MHz,
CDCl3): 7.11 (s, 2H, H at 5- and 6-position), 6.19 (s, 1H, H at 3-
position), 5.32 (s, 2H, CH3OCH2Oe), 3.55 (s, 3H, CH3OCH2Oe), 2.41
(s, 3H, eCH3 at 4-position).
55
l
n
n
a
n
a
o
a
r
t
X
X
X
n
o
M
M
C
D
Fig. 3. Comparison of the protective effects of DMXan, MXan and Xan on DNA against
the oxidation induced by Cu2þ/GSH at 150 min, and by AAPH at 360 min.
Compound 5 was obtained, yield 80%. 1H NMR (300 MHz, CDCl3):
9.90 (s, 1H, eCHO), 7.29 (d, 1H, J ¼ 9.0 Hz, H at 5-position), 6.80 (d,
1H, J ¼ 9.0 Hz, H at 6-position), 6.21 (s, 1H, H at 3-position), 4.75 (s,
2H, eOCH2CHO), 4.04 (s, 3H, eOCH3), 2.41 (s, 3H, eCH3 at 4-
position).
6.3.3. 4-Methyl-7-hydroxy-8-benzyloxycoumarin (9)
Benzylbromide (75 mmol) and K2CO3 (13.8 g, 100 mmol) were
dissolved in 150 mL of DMF, to which compound 8 (11.8 mg,
50 mmol) was added. The mixture was stirred for 6 h at room
temperature, then CHCl3 and water were added. The aqueous phase
was extracted by CHCl3, and the combined organic phase was
washed by water for three times, and dried over Na2SO4. After the
organic solvent was removed, the crude product was purified by
flash chromatography column with petroleum ether and ethyl
acetate (3:1, v:v) as eluent to obtain 4-methyl-7-methoxymethoxy-
8-benzyloxycoumarin, yield 95%. Then, 4-methyl-7-methox-
ymethoxy-8-benzyloxycoumarin (2.77 g, 8.5 mmol) was dissolved
in 250 mL of 2 N CH3COOH aqueous solution, and heated at 90 ꢂC
for 40 h. After the mixture was cooled to room temperature, pH of
the mixture was adjusted to 7 by adding NaHCO3. Then, the mixture
was extracted by ethyl acetate (50 mL ꢁ 4). The organic phase was
washed by brine, and dried over Na2SO4. After the organic solvent
was evaporated, compound 9 was obtained as a yellow solid
powder, yield 90%. 1H NMR (300 MHz, CDCl3): 7.46e7.37 (m, 5H, H
at Ph in PhCH2Oe), 7.25 (d, 1H, J ¼ 9.9 Hz, H at 5-positon), 6.87 (d,
1H, J ¼ 8.7 Hz, H at 6-positon), 6.16 (s, 1H, H at 3-positon), 6.07 (s,
1H, eOH), 5.30 (s, 2H, PhCH2Oe), 2.41 (s, 3H, eCH3 at 4-position).
6.2.5. 4-Methyl-8-methoxypsoralen (6)
Compound 5 (0.496 g, 2 mmol) was refluxed in 50 mL of 1 N
NaOH aqueous solution for 80 min under nitrogen. After the
mixture was cooled and acidified by 1 N HCl aqueous solution, the
crude product was precipitated, and purified by flash chromatog-
raphy column with petroleum ether and ethyl acetate (2:1, v:v) as
eluent to obtain compound 6, yield 50%. 1H NMR (300 MHz, CDCl3):
7.67 (d, 1H, J ¼ 1.2 Hz, H at 10-position), 7.48 (s, 1H, H at 5-position),
6.82 (d, 1H, J ¼ 1.5 Hz, H at 9-position), 6.25 (s, 1H, H at 3-position),
4.28 (s, 3H, eOCH3), 2.48 (s, 3H, eCH3 at 4-position).
6.2.6. 4-Methyl-8-hydroxylpsoralen (MXan)
Compound 5 (0.23 g, 1 mmol) was dissolved in 50 mL of CH2Cl2
and cooled to ꢃ10 ꢂC, to which 0.48 mL of 4.2 N CH2Cl2 solution of
BBr3 was added dropwisely. The mixture was stirred for 12 h at 0 ꢂC.
Then, 50 mL of water was added dropwisely, and stirred vigorously
for 1 h. The precipitate was purified by flash chromatography
column with petroleum ether and ethyl acetate (1:1, v:v) as eluent.
MXan was obtained and yield 93%. m. p. >250 ꢂC. 1H NMR
(300 MHz, DMSO-d6): 10.58 (s,1H, eOH), 8.07 (d, 1H, J ¼ 2.1 Hz, H at
10-position), 7.52 (s, 1H, H at 5-position), 7.03 (d, 1H, J ¼ 2.1 Hz, H at
9-position), 6.34 (s, 1H, H at 3-position), 2.48 (s, 3H, -CH3 at 4-
position); 13C NMR (DMSO-d6, 75 MHz): 159.4, 153.5, 146.9, 145.1,
139.4, 129.8, 124.6, 116.8, 112.2, 106.8, 106.6, 18.3. MS: m/z 217.7
(Mþ þ 1).
6.3.4. 4-Methyl-7-acetylmethoxy-8-benzyloxycoumarin (10)
Compound 10 was prepared by using compound 9 (1.13 g,
4 mmol) and bromoacetone (0.46 mL, 6 mmol) as reagents, and
following the procedure of the synthesis of compound 7, yield
86%. 1H NMR (300 MHz, CDCl3): 7.54 (d, 1H, J ¼ 6.3 Hz, H at 5-
position), 7.37e7.27 (m, 5H, H at Ph in PhCH2Oe), 6.75 (d, 1H,
Scheme 5. Methyl as prooxidative group, while hydroxyl group as antioxidative group in xanthotoxol.

C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
2565
J ¼ 9.0 Hz, H at 6-position), 6.19 (s, 1 H, H at 3-position), 5.23
(s, 2H, PhCH2Oe), 4.63 (s, 2H, CH3COCH2Oe), 2.40 (s, 3H,
CH3COCH2Oe), 2.23 (s, 3 H, eCH3 at 4-position).
dispensed into test tubes, each one contained 2.0 mL. The test tubes
were incubated in a water bath (37 ꢂC) to initiate the oxidation of
DNA. Three tubes were taken out at every 30 min and cooled
immediately, to which 1.0 mL of PBS1 solution of EDTA (30.0 mM as
the final concentration) was added to chelate Cu2þ. The tubes were
heated in boiling water for 30 min after 1.0 mL of thiobarbituric acid
(TBA) solution (1.00 g TBA and 0.40 g NaOH dissolved in 100 mL of
PBS1) and 1.0 mL of 3.0% trichloroacetic acid aqueous solution were
added. After the mixture in the test tube was cooled to room
temperature, 1.5 mL of n-butanol was added and shaken vigorously
to extract TBA reactive substance (TBARS). The absorbance of n-
butanol layer was measured at 535 nm, and plotted versus incu-
bation time.
6.3.5. 8-Benzyloxy-4,9-dimethylpsoralen (11)
Compound 10 (0.68 g, 2 mmol) was refluxed in 100 mL of 1 N
NaOH aqueous solution for 2 h under nitrogen. Then, the mixture
was cooled and acidified by 1 N HCl aqueous solution. The
precipitate was purified by flash chromatography column with
petroleum ether and ethyl acetate (2:1, v:v) as eluent. Compound 11
was obtained and yield 72%. 1H NMR (300 MHz, CDCl3): 7.54 (s, 1H,
H at 10-position), 7.46 (s,1H, H at 5-position), 7.35e7.26 (m, 5H, H at
Ph in PhCH2Oe), 6.25 (s,1H, H at 3-position), 5.52 (s, 2H, PhCH2Oe),
2.49 (s, 3H, eCH3 at 9-position), 2.26 (s, 3 H, eCH3 at 4-position).
6.6. Effects of DMXan, MXan and Xan on AAPH-induced oxidation
of DNA
6.3.6. 4,9-Dimethyl-8-hydroxylpsoralen (DMXan)
Compound 11 (0.32 g, 1 mmol) and 5% Pd/C (29 mg) were sus-
pended in 30 mL of ethyl acetate, and stirred overnight under H2
atmosphere. The mixture was filtered to remove catalyst, and the
organic phase was evaporated. The crude product was purified by
flash chromatography column with petroleum ether and ethyl
acetate (1:1, v:v) as eluent toobtain DMXan, yield 90%. m. p. >250 ꢂC.
1H NMR (300 MHz, DMSO-d6): 10.51 (s,1H, eOH), 7.85 (s,1H, H at 10-
position), 7.44 (s, 1H, H at 5-position), 6.33 (s, 1H, H at 3-position),
The oxidation of DNA induced by AAPH was carried out
according to the reference [20] with a little modification. Briefly,
the phosphate buffered solutions (PBS2: 8.1 mM Na2HPO4, 1.9 mM
NaH2PO4, 10.0 mM EDTA) of DNA and AAPH were mixed with DMSO
solutions of DMXan, MXan and Xan, respectively, in which the final
concentration for DNA, AAPH and DMXan, MXan or Xan were
2.0 mg/mL, 40.0 mM, and 80.0
mM, respectively. The following
2.50 (s, 3H, eCH3 at 9-position), 2.24 (s, 3H, eCH3 at 4-position). 13
C
operation was the same as that in the oxidation of DNA mediated by
Cu2þ/GSH except that the heating period was 15 min after TBA and
trichloroacetic acid aqueous solution were added to the mixture.
NMR (DMSO-d6, 75 MHz): 159.5, 153.6, 145.2, 143.0, 139.6, 129.7,
126.1, 116.5, 115.5, 112.0, 105.0, 18.4, 7.0. MS: m/z 231.7 (Mþ þ 1).
6.4. Antioxidant effectiveness of DMXan, MXan and Xan in chemical
experimental systems
6.7. Statistical analysis
All the data were the average values from at least three inde-
pendent measurements with the experimental error within 10%.
The data were analyzed by one-way ANOVA on Origin 6.0 profes-
sional Software, and p < 0.001 indicated a significance difference.
An emulsion was prepared by mixing 5.0 mg of
b-carotene,
40 mg of LH and 400 mg of Triton X-100 in 100 mL of double
distilled water under ultrasonic vibration [21]. The ethanol solu-
tions of DMXan, MXan and Xan (0.04 mL) were mixed with 2.96 mL
of
b
-carotene-LH emulsion, of which the final concentration of
M. The absorbance of the mixture
Acknowledgment
DMXan, MXan and Xan was 13.3
m
was measured at 460 nm, and plotted versus time.
The experiments of DMXan, MXan and Xan to trap ABTSþ
,
We thank the National Natural Science Foundation, China, for
the financial support.
ꢀ
DPPH, and galvinoxyl radical were carried out following the
description in literatures [22,24]. Briefly, ABTS (2.00 mL, 4.0 mM)
was oxidized by 1.41 mM K2S2O8 for 16 h to generate ABTSþꢀ, to
which 100 mL of ethanol was added to make the absorbance
(Absref) around 0.70 at 734 nm. DPPH and galvinoxyl radical were
dissolved in ethanol to make the absorbance (Absref) around 1.00 at
517 nm and 428 nm, respectively, to which ethanol solution of
References
[1] S. Ucar, A.R. Ozkan, Bioresour. Technol. 99 (2008) 8771e8776.
[2] N.H. Zawia, D.K. Lahiri, F. Cardozo-Pelaez, Free Radic. Biol. Med. 46 (2009)
1241e1249.
[3] K. Myung, J.A. Manthey, J.A. Narciso, Appl. Microbiol. Biotechnol. 78 (2008)
343e349.
DMXan, MXan or Xan was added with 200
mM as the final
[4] H. Kim, B.-S. Park, K.-G. Lee, C.Y. Choi, S.S. Jang, Y.-H. Kim, S.-E. Lee, J. Agric.
Food Chem. 53 (2005) 8537e8541.
concentration in the radical solutions. The decay of the absorbance
for DPPH and galvinoxyl radical solution (Absdetect) was recorded.
[5] J. Guo, X. Weng, H. Wu, Q. Li, K. Bi, Food Chem. 91 (2005) 287e292.
[6] J. Yu, I. Ghiviriga, B.S. Buslig, P. Cancalon, LWT Food Sci. Technol. 41 (2008)
420e424.
[7] J.-N. Cai, P. Basnet, Z.-T. Wang, K. Komatsu, L.-S. Xu, T. Tani, J. Nat. Prod. 63
(2000) 485e488.
[8] Z. Liu, Y. Lu, M. Lebwohl, H. Wei, Free Radic. Biol. Med. 27 (1999) 127e133.
[9] J. Grillari, H. Katinger, R. Voglauer, Nucleic Acids Res. 35 (2007) 7566e7576.
[10] J.J. Serrano-Pérez, L. Serrano-Andrés, M. Merchán, Chem. Phys. 347 (2008)
422e435.
[11] B. Girennavar, G.K. Jayaprakasha, Y. Jadegoud, G.A.N. Gowda, B.S. Patil, Bioorg.
Med. Chem. 15 (2007) 3684e3691.
The same operation was performed to ABTSþ with 40
mM as the
ꢀ
final concentration for DMXan, MXan or Xan. The percentages of
these radicals trapped by DMXan, MXan and Xan were calculated
by (1ꢃAbsdetect/Absref) ꢁ 100, and plotted versus incubation time.
6.5. Effects of DMXan, MXan and Xan on the oxidation of DNA
mediated by Cu2þ/GSH
[12] O.M.A. Hafez, K.M. Amin, N.A. Abdel-Latif, T.K. Mohamed, E.Y. Ahmed,
T. Maher, Eur. J. Med. Chem. 44 (2009) 2967e2974.
Cu2þ/GSH-mediated oxidation of DNA was carried out according
to the reference [19] with a little modification. Briefly, CuSO4
aqueous solution was mixed with the phosphate buffered solutions
(PBS1: 6.1 mM Na2HPO4, 3.9 mM NaH2PO4) of DNA and GSH, in
which the final concentration for DNA, Cu2þ and GSH were 2.0 mg/
mL, 5.0 mM, and 3.0 mM, respectively. Then, dimethyl sulfoxide
(DMSO) solution of DMXan, MXan or Xan was added with the final
concentration at 0.2 mM, respectively. The above mixture was
[13] Z.-Q. Liu, W. Yu, Z.-L. Liu, Chem. Phys. Lipids 103 (1999) 125e135.
[14] W. He, B.-L. Zhang, Synth. Commun. 37 (2007) 361e367.
[15] T. Finkel, N.J. Holbrook, Nature 408 (2000) 239e247.
[16] S. Mena, A. Ortega, J.M. Estrela, Mutat. Res. 674 (2009) 36e44.
[17] M.S. Cooke, M.D. Evans, M. Dizdaroglu, J. Lunec, FASEB J. 17 (2003) 1195e1214.
[18] R.S. Lankalapalli, J.T. Eckelkamp, D. Sircar, D.A. Ford, P.V. Subbaiah, R. Bittman,
Org. Lett. 11 (2009) 2784e2787.
[19] C.J. Reed, K.T. Douglas, Biochem. J. 275 (1991) 601e608.
[20] P. Zhang, S.T. Omaye, Food Chem. Toxicol. 39 (2001) 239e246.

2566
C. Xiao et al. / European Journal of Medicinal Chemistry 45 (2010) 2559e2566
[21] J. Yu, L. Wang, R.L. Walzem, E.G. Miller, L.M. Pike, B.S. Patil, J. Agric. Food Chem.
53 (2005) 2009e2014.
[22] H. Shi, E. Niki, Lipids 33 (1998) 365e370.
[23] P. Bovicelli, J. Pharm. Pharmacol. 59 (2007) 1703e1710.
[24] Y.-Z. Tang, Z.-Q. Liu, J. Am. Oil Chem. Soc. 84 (2007) 1095e1100.
[25] S. Oikawa, K. Murakami, S. Kawanishi, Oncogene 22 (2003) 3530e3538.
[26] W.A. Prütz, Biochem. J. 302 (1994) 373e382.
[27] V.W. Bowry, R. Stocker, J. Am. Chem. Soc. 115 (1993) 6029e6044.
[28] Y. Sato, S. Kamo, T. Takahashi, Y. Suzuki, Biochemistry 34 (1995)
8940e8949.
[29] A. Okamoto, K. Tanabe, I. Saito, Org. Lett. 3 (2001) 925e927.
[30] A. Russell, J.R. Frye, Org. Syn. 3 (1955) 281e284.