Journal of Ethnopharmacology 87 (2003) 73–83

Isolation and characterization of antioxidant phenolic

compounds from the aerial parts of Hypericum hyssopifolium
L. by activity-guided fractionation
Ahmet Cakir a , Ahmet Mavi a , Ali Yıldırım a,∗ , Mehmet Emin Duru b ,
Mansur Harmandar b , Cavit Kazaz c
a Atatürk Üniversitesi, Kazım Karabekir Eğitim Fakültesi, Kimya Eğitimi Anabilim Dalı, 25240 Erzurum, Turkey
b Muğla Üniversitesi, Fen-Edebiyat Fakültesi, Kimya Bölümü, 48000 Muğla, Turkey
c Atatürk Üniversitesi, Fen-Edebiyat Fakültesi, Kimya Bölümü, 25240 Erzurum, Turkey

Received 7 August 2002; received in revised form 12 March 2003; accepted 21 March 2003

Abstract

Dried methanol extract of Hypericum hyssopifolium subsp. elongatum var. elongatum was dissolved in distilled water, and then fractioned
by re-extracting with petroleum ether, chloroform, and ethyl acetate, subsequently. Antioxidant and 1,1-diphenyl-2-picrylhydrazyl (DPPH)
radical-scavenging activities of these fractions were determined, in vitro. The amounts of total phenolic compounds were also determined.
None of these fractions showed antioxidant activity, in contrast water and ethyl acetate fractions acted as prooxidant. However, the ethyl acetate
fraction exhibited the highest DPPH radical-scavenging activity and the amount of its total phenolic compound was highest, too. Therefore,
ethyl acetate fraction was subjected to further separation by chromatographic methods. Thus, five flavonoids (I3,II8-biapigenin, quercetin,
quercetin-3-O-␣-l-arabinofuranoside, quercetin-3-O-␤-d-galactopyranoside, quercetin-3-O-␤-d-galactopyranoside-7-O-␤-d-glucopyrano-
side) and a napthodianthrone (hypericin) were isolated, and their structures were determined by UV, IR, NMR, and MS spectroscopic
methods. All isolated compounds showed antioxidant and DPPH radical-scavenging activities. Although, I3,II8-biapigenin and hypericin
were able to show highest antioxidant activity, they had the lowest DPPH radical-scavenging activities. From these results, it can be suggested
that these compounds may be used as potential antioxidants. In addition, the petroleum ether fraction was subjected to silica gel column
chromatography (CC). Then, n-dotriacontanyl hexadecanoate, bis(2-methylheptyl) phthalate, and ␤-sitosterol were isolated from it. It is of
interest to present the spectral data of bis(2-methylheptyl) phthalate first time in the present study.
© 2003 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Guttiferae; Hypericum hyssopifolium; Phthalic acid ester; Flavonoids; Antioxidant activity; Radical scavenging

1. Introduction antidepressant, anxiety, antiviral, wound healing and an-

Hypericum (Guttiferae) is a large genus of herbs or
shrubs, which grows widely at temperate regions, and used
as traditional medicinal plants in various parts of the world
(Yazaki and Okada, 1994). In recent years, the antide-
pressant activity of Hypericum perforatum L., known as
St. John’s wort, caused to the widespread interest in the
study of the Hypericum genus (Hu and Sim, 2000). Several
studies have been published concerning Hypericum perfo-
ratum L. activities, which have shown that this species has
∗ Corresponding author. Tel.: +90-442-2314020;
fax: +90-442-2360955.
E-mail addresses: ayildirim61@yahoo.com, ayil@atauni.edu.tr
(A. Yıldırım).
timicrobial activities (Barnes et al., 2001; Butterweck et al.,
1997, 2000, 2002; Flausino et al., 2002; Sakar and Tamer,
1990). For example, it has been reported that Hypericum
perforatum extract LI 160 is a potentially effective for treat-
ment of children less than 12 years with symptoms related
to depression and psychovegetative disorders (Hubner and
Kirste, 2001). Therefore, Hypericum genus has attracted
much attention in investigation metabolites from this
genus, many of which are biologically active compounds
with phloroglucinol moiety (Decosterd et al., 1989, 1991;
Jayasuriya et al., 1989; Ishiguro et al., 1990; Sanchez-Mateo
et al., 2002; Strolin-Benedetti and Dosterd, 1992). It has
been reported that Hypericum genus contained some an-
tiviral prenylated pholoroglucinol derivatives (Tada et al.,
1991). The methanol extracts as well as the chloroform and

0378-8741/03/$ – see front matter © 2003 Elsevier Science Ireland Ltd. All rights reserved.
doi:10.1016/S0378-8741(03)00112-0
74 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
butanol fractions obtained from the aerial parts of Hyper-
icum reflexum were reported to possess good antibacterial
properties against the Gram-positive bacteria (Herrera et al.,
1996). Hypericum species contains also the variety of com-
pounds such as flavonoids (Chung et al., 1997; Dias et al.,
1998; Ishiguro et al., 1991, 1993a,b), xanthones (Gunatilaka
et al., 1979; Rath et al., 1996; Wu et al., 1998a), chromenyl
ketones (Ishiguro et al., 1990; Wu et al., 1998b), hyper-
forin derivatives (Decosterd et al., 1989; Maisenbacher
and Kovar, 1992; Trifunovic et al., 1998), phloroglucinols
(Decosterd et al., 1991; Ishiguro et al., 1998; Ishiguro et al.,
1994), n-alkanes (Brondz et al., 1983), napthodianthrones
(Kitanov, 2001) and essential oils (Cakir et al., 1997).
Hypericin, a powerful naturally occurring photosensitizer,
is found in Hypericum species. In recent years, increased
interest in hypericin as a potential clinical anticancer agent
has arisen for several studies established its powerful in vivo
and in vitro antineoplastic activity upon irridation (Agostinis
et al., 2002; Vantieghem et al., 2001). It has been also
found that napthodianthrones, hypericin and pseudohyper-
icin, posses high antiretroviral activity in vitro and in vivo,
suggested that these compounds may be interesting in the
search of new tools against the acquired immune disease
syndrome (Meruelo et al., 1988) and also antidepressant ac-
tivity (Butterweck et al., 2002).
Lipid peroxidation is one of the major factors causing
deterioration of foods during the storage and processing.
Oxidized polyunsaturated fatty acids may induce aging and
carcinogenesis. Although there are some synthetic antioxi-
dant compounds, such as butylated hydroxytoluene (BHT)
and butylated hydroxyanisole (BHA), which are commonly
used in processed foods, it has been reported that these com-
pounds have some side effects (Branien, 1975; Ito et al.,
1983). Therefore, many researchers have focused on natural
antioxidant sources.
There are about 80 species of Hypericum genus in Turkey
and these species have been used for a long time for the
treatment of external wound and gastric ulcer and also as
sedative, antiseptic, and antispazmatic in folk medicine
(Baytop, 1984). There is so far no report related to Hyper-
icum hyssopifolium L.
The aim of present study was to determine the antioxidant
potential of Hypericum hyssopifolium and its phenolic sec-
ondary metabolites. Therefore, the antioxidant and DPPH
radical-scavenging activities of the petroleum ether, chloro-
form, ethyl acetate, and water fractions of methanol extract
were determined. As ethyl acetate fraction was able to
show the highest DPPH radical-scavenging activity and the
amount of phenolic compound was highest on it, this frac-
tion was subjected to further separation by chromatographic
methods. Thus, five flavonoids and one napthodianthrone
were isolated from this fraction and their antioxidant and
DPPH radical-scavenging activities were determined. In ad-
dition, n-dotriacontanyl hexadecanoate, bis(2-methylheptyl)
phthalate, and ␤-sitosterol, which are not phenolic com-
pounds, were isolated from petroleum ether fraction. Spec-
tral data of bis(2-methylheptyl) phthalate isolated from this
fraction were presented for the first time.
2. Methods
2.1. Plant material
Hypericum hyssopifolium subsp. elongatum var. elon-
gatum, authenticated by Dr. Yusuf Kaya, Department of
Botany, Faculty of Science, Ataturk University, was col-
lected at flowering period in June 1998 in Erzurum. A
voucher specimen has been kept in Department of Chem-
istry, K.K. Education Faculty, Atatürk University, Erzurum
(Turkey).
2.2. Analytical material and methods
Melting points were determined on a Reicher-Jung
Kofler apparatus. NMR spectra were recorded on a Varian
200 MHz spectrometer, operating at 200 and 50 MHz for
1 H and 13 C, respectively using deuterochloroform (CDCl ),
3
dimethylsulfoxide-d6 (DMSO-d6 ), and deuteromethanol
(CD3 OD). Chemical shifts are expressed in δ (ppm) down-
field from tetramethylsilane (TMS) as an internal standard
and coupling constants reported in Hz. UV spectra were
recorded on a JASCO V-530 spectrometer. The IR spectra
were determined on a Shimadzu FT-IR 8000 spectropho-
tometer. Electron impact-mass spectroscopy (EI-MS) and
high resolution-mass spectroscopy (HR-MS) were mea-
sured on VG zabspect; chromatron Model 7924T. Column
chromatography (CC) was carried out using silica gel
60 (70–230 mesh), thin layer chromatography (TLC) and
prep. TLC on silica gel 60 precoated plates, F-254 (Merck).
The spots on TLC were visualized by spraying with 1%
vanilin-H2 SO4 followed by heating and hold to vapour of
NH3 .
2.3. Extraction procedures
The dried aerial parts of Hypericum hyssopifolium (800 g)
were extracted with methanol (MeOH) (3 × 4 l). The MeOH
extract was evaporated under vacuum to dryness as a dark
brown mass (85.4 g) and then the concentrated MeOH
extract was dissolved in distilled H2 O. The solution suc-
cessively was partitioned with petroleum ether (40–60◦ ),
CHCl3 , and ethyl acetate (EtOAc), yield with 6.9, 2.9, and
12.5 g, respectively. Then, water fraction was lyophilized in
a Labconco 117 freeze-dryer at 5 ␮mHg and −50 ◦ C.
2.4. Isolation procedures
The concentrated EtOAc fraction (7 g) was fractioned
on silica gel CC (200 g, 70–230 mesh) using petroleum
ether–EtOAc (1:1, 1:3, and 0:1) and EtOAc–MeOH (3:1,
1:1, 1:3, and 0:1 v/v) giving seven fractions (A-G). Fraction
 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
Scheme 1.
A was determined to contain only compound 2 (25 mg),
monitoring by TLC. Fraction B (156 mg), eluted with
EtOAc–petroleum ether (3:1), was further subjected to sil-
ica gel CC (50 g, 70–230 mesh) with EtOAc–petroleum
ether (3:1) and compound 1 (75 mg) and compound 2
(24 mg) were purified. Fraction C (1.9 g) was re-portioned
to two fractions (fractions 1 and 2) on silica gel CC (100 g)
eluting with EtOAc–MeOH (9:1). Fraction 1 (955 mg)
was further separated on CC using 75 g of silica gel and
EtOAc–formic acid–acetic acid–MeOH (100:1:1:5) and
compound 3 (200 mg) and compound 4 (43 mg) were iso-
lated. Fraction D (1.92 g) was subjected to chromatography
on a silica gel (100 g) column eluting with EtOAc–MeOH
(4:1) and yielded compound 5 (310 mg) and compound 6
(44 mg), Scheme 1.
The concentrated petroleum ether fraction (6.8 g) was
subjected to a silica gel CC (180 g, 70–230 mesh) using
petroleum ether–EtOAc (gradient, 1:0 → 0:1) solvent mix-
75
tures and obtained six fractions (1–6). Fraction 2 (1.3 g),
eluted from the initial CC, was further separated by sil-
ica gel CC (100 g, 70–230 mesh) using n-hexane–EtOAc
with increasing polarity (9:1 → 3:7) and obtained six frac-
tions. Compound 7 (95 mg) was isolated from the fraction
1 (155 mg) over silica gel CC using n-hexane–CHCl3 (9:1)
followed by crystallization in n-hexane. Elution of fraction
3 (155 mg) with n-hexane–CHCl3 (4:1) on a silica gel CC
yielded compound 8 (25 mg). Fractions D and F were fur-
ther chromatographed on silica gel (n-hexane–Et2 O; 3:1)
and EtOAc–petroleum ether; (7:3), respectively, and com-
pounds 9 and 1 (20 mg and 85 mg, respectively) were ob-
tained, Scheme 2.
2.5. Antioxidant activity
Antioxidant activities of the petroleum ether, chloroform,
ethyl acetate, and water fractions of methanol extract, and
76 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
Scheme 2.
pure phenolic compounds isolated from ethyl acetate frac-
tion were measured.
The antioxidant activity was determined according to
the thiocyanate method. Briefly, 150 ␮l fraction solu-
tion (1 ␮g ml−1 ) or 500 ␮l isolated compound solution
(1 ␮g ml−1 ) was mixed with 2.5 ml of 0.02 M linoleic acid
(Fluka) emulsion (contains equal weight of Tween-20 in pH
7.4 phosphate buffered saline, Sigma) and the final volume
was adjusted to 5 ml with phosphate buffered saline (pH 7.4)
in a test tube and incubated in darkness at 37 ◦ C. The amount
of peroxide was determined by measuring absorbance at
500 nm after coloring with 0.1 ml FeCl2 (0.02 M) and 0.1 ml
thiocyanate (30%) at intervals during incubation (Yen and
Chen, 1995; Yildirim et al., 2001). To eliminate the sol-
vent effect, control test sample, which contains the same
amount of solvent added into the linoleic acid emul-
sion in the tests of fractions and pure compounds, was
used.
2.6. DPPH radical-scavenging activity
DPPH radical-scavenging activities of the petroleum
ether, chloroform, ethyl acetate, and water fractions of
methanol extract were measured. Then, fractions of ethyl
acetate whose activity was the highest one was divided to
seven fractions by CC and DPPH radical-scavenging ac-
tivities of each fraction were compared. The activities of
pure compounds isolated from these fractions were also
determined.
Experiments were carried out according to the method
of Blois (1958) with a slight modification. Briefly, a 1 mM
solution of DPPH (Fluka) radical solution in methanol was
prepared and then, 1 ml of this solution was mixed with
3 ml of sample solutions in ethanol. Finally, after 30 min,
the absorbance was measured at 517 nm. Decreasing of
the DPPH solution absorbance indicates an increase of the
DPPH radical-scavenging activity. This activity is given
as % DPPH radical-scavenging that is calculated in the
equation:
% DPPH radical-scavenging


control absorbance − sample absorbance
= × 100
control absorbance
The DPPH solution without sample solution was used as
control.
2.7. Amount of total phenolic compounds
Antioxidant compounds generally contains phenolic
group(s). Because of this, amounts of phenolic compounds
in each of the fractions of methanol extract were compared
to obtain more information about the fraction(s) which
posses(s) antioxidant potential.
This was carried out as described previously (Singleton
et al., 1999). Briefly, extract solution was transferred in to
100-ml Erlenmeyer flask then final volume was adjusted
to 46 ml by addition of distilled water. Afterward, 1 ml of
Folin-Ciocalteu Reactive (FCR) (Fluka) was added into this
mixture and after 3 min 3 ml of Na2 CO3 (2%) was added.
Subsequently, mixture was shaken on a shaker for 2 h at
 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83 77

Fig. 1. Antioxidant activities of methanol fractions and BHT. Indicated amounts of samples (␮g) were added into linoleic acid emulsion. The results are
means of two different experiments in each of which three measurements were made. The data points without error bar indicate that standard deviation
is too small to be seen.

room temperature and then absorbance was measured at 2.9. The spectral information of bis(2-methylheptyl)
760 nm. phthalate

2.8. Statistical analyses of results of activity studies
In the measurements of the activities, two independent ex-
periments in each of which three measurements were made,
and results were calculated as means of these measurements.
Using SPSS 9.0 software statistical calculations were car-
ried out. To determine whether there was any differences
between activities of samples, variance analysis was applied
to the results. Values of P < 0.05 were considered as sig-
nificantly different (α = 0.05).
UV λmax (log ε) (CHCl3 ) 250 (1.73), 276 (1.61) nm;
IR ν max (KBr): 2960–2860 (C-H), 1724 (␣,␤-unsaturated
C==O), 1464, 1380, 1277, 1124, 1074, 908, 735 cm−1 ; 1 H
NMR, (200 MHz, CDCl3 ) ppm δ: 7.74–7.57 (4 aromatic H,
m, AA′ BB′ , H-3 and H-4), δ: 4.26 (2H, dd, J1 = 8.5 Hz,
J2 = 1.6 Hz, H-1′ a), δ: 4.19 (2H, dd, J1 = 8.5 Hz, J2 =
1.6 Hz, H-1′ b), δ: 1.68 (2H, m, H-2′ ), δ: 1.21-1.50 (16H,
m, H-3′ , H-4′ , H-5′ and H-6′ ), δ: 0.93 (6H, t, J = 7.4 Hz,
H-7′ ), δ: 0.88 (6H, t, J = 6.2 Hz, H-8′ ). 13 C NMR, (50 MHz,
CDCl3 ) ppm δ: 169.7 (C-1), δ: 134.5 (C-2), ␦: 132.8 (C-4),

Fig. 2. Comparison of DPPH radical-scavenging activities of methanol fractions. Indicated amounts of fractions (␮g) were added into the DPPH solution
(1 mM, 1 ml). The results are means of two different experiments, in each of which three measurements were made. The data point without error bar
indicates that standard deviation is too small to be seen.
78 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83

Fig. 3. Comparison of amount of total phenolic compound of methanol fractions (500 ␮g). The results are means of two different experiments, in each
of which three measurements were made. The data points without error bar indicate that standard deviation is too small to be seen.

δ: 130.8 (C-3), δ: 70.2 (C-1′ ), δ: 40.7 (C-2′ ), δ: 32.4 (C-3′ ), 3. Results

δ: 30.9 (C-4′ ), δ: 25.8 (C-5′ ), δ: 25.0 (C-6′ ), δ: 16.0 (C-8′ ),
δ: 12.9 (C-7′ ). EI-MS 70 eV, m/z (rel. int.): 390 [M]+ (2),
279 [M-C8 H17 ; C16 H22 O4 ]+ (74), 167 [C8 H6 O4 ; phthalic
acid]+ (85), 149 [C8 H5 O3 ]+ (100), 113 [C8 H17 ]+ (58), 104
[C7 H5 O]+ (21), 83 [C6 H13 ]+ (37), 71 [C5 H11 ]+ (68), 57
[C4 H9 ]+ (73). HR-MS: m/z 390.27710, calc. for C24 H38 O4
390.275415.
The dried aerial parts of Hypericum hyssopifolium were
extracted with methanol for three times. The concentrated
methanol extract was then dissolved in a small amount of
distilled water. Afterwards, it was extracted with petroleum
ether, chloroform, and ethyl acetate, successively, and then
water was removed by lyophilization. Antioxidant activities

Fig. 4. DPPH radical-scavenging activities of ethyl acetate fractions. Indicated amounts of fractions (␮g) were added into the DPPH solution (1 mM,
1 ml), and volumes were adjusted to 4 ml adding ethanol. The results are means of two different experiments, in each of which three measurements were
made. The data points without error bar indicate that standard deviation is too small to be seen.
 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83 79

of each of fractions (50, 100, and 150 mg) were measured.
At the all studied concentrations there were prooxidant ac-
tivities in water and ethyl acetate fractions (P < 0.05 versus
control). However, the other fractions were not statistically
different from control (P > 0.05). For the sake of simplifi-
cation, only the results of samples containing 150 mg frac-
tions were given in Fig. 1.
DPPH radical-scavenging activities of each of fractions
(100, 200, 400 mg) were measured. The highest DPPH
radical-scavenging activity was shown by ethyl acetate frac-
tion and the lowest one was chloroform fraction (Fig. 2).
As absorbance values were too low at lower concentra-
tions, 500 ␮g of fractions were used to determine the amount
of total phenolic compounds. The ethyl acetate fraction had
the highest amount of total phenolic compound (Fig. 3).
The ethyl acetate fraction showed the highest DPPH
radical-scavenging activity and amount of total pheno-
lic compounds. Thus it was further fractioned into seven
fractions (fractions A–G) on silica gel CC eluting with
petroleum ether–ethyl acetate (1:1, 1:3, 0:1 v/v) and ethyl
acetate–methanol (3:1, 1:1, 1:3 v/v). All fractions were
controlled by TLC and tested for DPPH radical-scavenging
activities. As can be seen from Fig. 4, all fractions (A–G)
exhibited DPPH radical-scavenging activity. In addition,
compounds 1–6 were isolated from the fractions (A–D),
which showed relatively high DPPH radical-scavenging
activity (Fig. 5).
The structures of these compounds were characterized by
UV, IR, MS, 1 H- and 13 C-NMR and 2D-NMR methods as
I3,II8-biapigenin, quercetin, quercetin-3-O-␣-l-arabinofur-
anoside, quercetin-3-O-␤-d-galactopyranoside, quercetin-
3-O-␤-d-galactopyranoside-7-O-␤-d-glucopyranoside, and
hypericin. Their structures were also confirmed by com-
paring the previously reported spectral data (Berghöfer and
Hölzl, 1987). Quercetin-3-O-␣-l-arabinofuranoside and qu-
ercetin-3-O-␤-d-galactopyranoside-7-O-␤-d-glucopyrano-
side were isolated from Hypericum species for the first time
in the present study.
DPPH radical-scavenging and antioxidant activities of
isolated compounds from ethyl acetate fraction were evalu-
ated. The results show that all compounds exhibited DPPH
radical-scavenging (Fig. 5) and antioxidant activities com-
pared to the control (P < 0.05) (Fig. 6a and b). Among
the isolated compounds, the highest activities were observed
for I3,II8-biapigenin and hypericin. However, the antioxi-
dant activities of isolated compounds were lower than that
of synthetic antioxidant, BHT.
Compounds 7–9 were isolated together with I3,II8-bia-
pigenin, which was also isolated from ethyl acetate solu-
ble fraction, using silica gel CC from the petroleum ether
fraction. These compounds were identified as n-dotria-
contanyl hexadecanoate, bis(2-methylheptyl) phthalate, and
␤-sitosterol by spectroscopic methods and comparing the
IR, 1 H- and 13 C-NMR and MS spectral information with
known compounds (Berghöfer and Hölzl, 1987; Lu and
Lin, 1994; and Rubenstein et al., 1976, respectively). In
the present study, the spectral data of bis(2-methylheptyl)
phthalate (8) were presented for the first time. As these

Fig. 5. Comparison DPPH radical-savenging activities of isolated compound from ethylacetate fractions and BHT. Compound 1: I3,II8-biapigenin,
compound 2: quercetin, compound 3: quercetin-3-O-␣-l-arabinofuranoside, compound 4: quercetin-3-O-␤-d-galactopyranoside, compound 5:
quercetin-3-O-␤-d-galactopyranoside-7-O-␤-d-glucopyranoside, compound 6: hypericine, BHT: buthylated hydroxytoluene. Fifty micrograms of com-
pounds were added into DPPH solution (1 mM, 1 ml), and volumes were adjusted to 4 ml adding ethanol. The results are means of two different
experiments, in each of which three measurements were made.
80 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83

Fig. 6. Comparison of antioxidant activities of BHT and isolated compounds from ethylacetate fractions. (a) Control: linoleic acid emul-
sion without sample, compound 1: I3,II8-biapigenin, compound 2: quercetin, compound 3: quercetin-3-O-␣-l-arabinofuranoside, BHT: buthy-
lated hydroxytoluene. (b) Control: linoleic acid emulsion without sample, compound 4: quercetin-3-O-␤-d-galactopyranoside, compound 5:
quercetin-3-O-␤-d-galactopyranoside-7-O-␤-d-glucopyranoside, compound 6: hypericine, BHT: buthylated hydroxytoluene. Five hundred micrograms of
BHT or isolated compounds were added into the linoleic acid emulsion (0.02 M, pH 7.0). The results are means of two different experiments, in each
of which three measurements were made. The data points without error bar indicate that standard deviation is too small to be seen.

compounds are not phenolic compounds and due to solubil-
ity difficulties in linoleic acid emulsion system, the activities
of these compounds were not studied.
The EI-mass spectrum of compound 8 isolated from the
petroleum ether fraction of methanol extract of Hypericum
hyssopifolium revealed significant fragment ions at m/z 390
[M]+ , 166 [phthalic acid]+ , 149 [C8 H5 O3 ]+ . This frag-
mentation ions indicated that this compound, which has the
molecular formula C24 H38 O4 [M]+ , must be a phthalate.
The presence of a phthalyl group in compound 8 was shown
by the IR bands at 1724, 1464, 1380, 1277, and 735 cm−1
and by the 1 H NMR signals at δ 7.57–7.74 (4 aromatic H,
m, AA′ BB′ system) and 13 C NMR signals at δ 169.7, 134.5,
132.8, and 130.8. The presence of only four carbon signals
for the phthalyl moiety clearly demonstrates that both the
carboxylic groups were esterified (Johnson and Jankowski,
1978; Shukla and Thakur, 1990). 1 H NMR spectrum of com-
pound 8 showed the presence of two AB systems with the
 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
signals at δ 4.19 (2H, dd, J1 = 8.5 Hz, J2 = 1.6 Hz) and
δ 4.26 (2H, dd, J1 = 8.5 Hz, J2 = 1.6 Hz) correspond-
ing to methylene protons adjacent to ester oxygen. This AB
system indicated the presence of a chiral group adjacent to
methylene ester and of symmetry in the molecule structure.
The methyl protons (12H) appeared as a t (J = 7.4 Hz) at
δ 0.93 and a d (J = 6.2 Hz) at δ 0.88. The signal of H-2′
methine protons (2H) was observed as an m centered at δ
1.68. The observation of only eight peaks at aliphatic region
in 13 C NMR spectrum was also evidence to the existence of
symmetry in the structure. While two methyl group signals
appeared at δ 16.0 and 12.9, ester methylene carbon sig-
nal appeared at δ 70.2. The correlation in 2D 1 H-13 C Het-
eronükleer COSY spectra and APT pulse sequence allowed
also us to assign of all the carbons and to confirm the struc-
ture. All these data supported the structure of compound 8
as bis(2-methylheptyl) phthalate.
4. Discussion
In the present study, antioxidant activities of petroleum
ether, chloroform, ethyl acetate, and water fractions of
methanol extract of Hypericum hyssopifolium and five
flavonoids and a napthodianthrone (which were isolated
from ethyl acetate fraction) were determined by using the
thiocyanate method. In this method, the amount of perox-
ides formed in emulsion during incubation is determined
spectrophotometrically by measuring absorbance at 500 nm.
High absorbance is an indication of high concentration of
formed peroxides. Therefore, low absorbance indicates high
antioxidant activity (Yen and Chen, 1995).
Although BHT that is known as a standard antioxidant
compound was able to inhibit the formation of peroxides,
none of the fractions showed antioxidant activity, Fig. 1. In
contrast, as can be seen from the figure, ethyl acetate and
water fractions acted as prooxidant. Thus, the absorbances
of test samples containing these fractions were higher than
control (P < 0.05).
DPPH radical-scavenging activities of above samples
were also determined. The ethyl acetate fraction showed
the highest activity (Fig. 2). This fraction also contained
the highest amount of total phenolic compounds (Fig. 3).
Both of them can be accepted an indication of antioxidant
potential (Yildirim et al., 2001). Therefore, this fraction
was divided into seven fractions and it was found that all of
these fractions were able to show DPPH radical-scavenging
activities (Fig. 4). Six phenolic compounds, which were
isolated from these fractions, were also able to show this
activity (Fig. 5).
Although ethyl acetate fraction of methanol extract did not
show antioxidant activity, all phenolic compounds isolated
from this fraction showed the activity. The highest antiox-
idant activity was shown by I3,II8-biapigenin, hypericine
and BHT. Differences between them were not statistically
significant. At first glance, it seems that there is a contradic-
81
tion between these results. However, ethyl acetate fraction
may contain several compounds. Some of them might act as
prooxidant while the others act as antioxidant. The sum of
these activities determines the decisive activity in the frac-
tion. Therefore, one can suggest that prooxidant activity is
dominant in this fraction. It was aimed to isolate antioxi-
dant compounds and determine their activities. These puri-
fied compounds showed antioxidant activity in the linoleic
acid emulsion system.
Ethyl acetate fraction was able to show high DPPH
radical-scavenging activity, while it did not show antioxi-
dant activity. However, I3,II8-biapigenin (1) and hypericin
(6) had the lowest DPPH radical-scavenging activities,
they had the highest antioxidant activities among the iso-
lated compounds. These results may suggest that although
DPPH radical-scavenging activity might be an indication
of potential antioxidant activity, there may not be always a
linear correlation between these two activities. The antioxi-
dant activities of putative antioxidants have been attributed
to various mechanisms; among these are prevention of
chain initiation, binding of transition metal ion catalysts,
decomposition of peroxides, prevention of continued hy-
drogen abstraction, and radical scavenging (Diplock, 1997;
Heinoman et al., 1998).
From the other side, solubility of a compound in the re-
action environment may affect its activity. In general, the
solubility of aglycones in emulsion phase is more than that
of glycosides, which naturally choose to move water phase
in an oil-in-water emulsion (Huang et al., 1996). High an-
tioxidant activities of I3,II8-biapigenin (1) and hypericin (6),
which are aglycones, may be attributed to their solubility in
oil phase. Lower antioxidant activities of glycosides may be
due to this solubility effect. In addition, it has been reported
that free –OH groups in phenolic compounds are mainly re-
sponsible for antioxidant activity (Weng and Wang, 2000).
I3,II8-biapigenin (1) and hypericin (6) contain more hy-
droxyl groups than that of other compounds isolated. These
may be also the cause of the higher antioxidant activities of
I3,II8-biapigenin (1) and hypericin (6).
Consequently, six phenolic compounds have been iso-
lated from ethyl acetate fraction of methanol extract of Hy-
pericum hyssopifolium. From the obtained results we could
suggest that these compounds can be used as antioxidant in
oil-in-water emulsion system.
References
Agostinis, P., Vantieghem, A., Merlevede, W., de Witte, P.A.M., 2002.
Hypericin in cancer treatment: more light on the way. International
Journal of Biochemistry and Cell Biology 34, 221–241.
Barnes, J., Anderson, L.A., Phillipson, J.D., 2001. St. John wort (Hy-
pericum perforatum L.): a review of its chemistry pharmacology
and clinical properties. Journal of Pharmacy and Pharmacology 53,
583–600.
Baytop, T., 1984. Therapy with Medicinal Plants in Turkey. Istanbul
University Publication No 3255, Istanbul, pp. 166–167.
82 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
Branien, A.L., 1975. Toxicology and biochemistry of butylated hydrox-
yanisole and butylated hydroxytoluene. Journal of The American Oil
Chemists Society 52, 59–63.
Berghöfer, R., Hölzl, J., 1987. Biflavonoids in Hypericum perforatum;
part 1. Isolation of I3,II8-biapigenin. Planta Medica 53, 216–217.
Blois, M.S., 1958. Antioxidant determinations by the use of a stable free
radical. Nature 181, 1198–1200.
Brondz, I., Greibrokk, J., Aasen, A.J., 1983. n-Alkanes of Hypericum
perforatum: a revision. Phytochemistry 2, 295–296.
Butterweck, V., Wall, A., Lieflander-Wulf, U., Winterhoff, H., Nahrsted,
A., 1997. Effects of the total extract and fractions of Hypericum perfo-
ratum in animal assays for antidepressant activity. Pharmacopsychiatry
30, 117–124.
Butterweck, V., Jurgenliemk, G., Nahrstedt, A., Winterhoff, H., 2000.
Flavonoids from Hypericum perforatum show antidepressant activity
in the forced swimming test. Planta Medica 66, 3–6.
Butterweck, V., Bockers, T., Korte, B., Wittkowski, W., Winterhoff, H.,
2002. Long-term effects of St. John’s wort and hypericin on monoamine
levels in rat hypothalamus and hippocampus. Brain Research 930, 21–
29.
Cakir, A., Duru, M.E., Harmandar, M., Ciriminna, R., Passannanti, S.,
Piozzi, F., 1997. Comparison of the volatile oils of Hypericum scabrum
L. and Hypericum perforatum L. from Turkey. Flavour and Fragrance
Journal 12, 285–287.
Chung, M.I., Lai, M.H., Yen, M.H., Wu, R.R., Lin, C.N., 1997. Phenolics
from Hypericum geminiflorum. Phytochemistry 44, 943–947.
Decosterd, L.A., Stoeckli-Evans, H., Chapuis, J.C., Msonthi, J.D., Sordat,
B., Hostettmann, K., 1989. New hyperforin derivatives from Hypericum
revolutum Vahl with growth-inhibitory activity against a human colon
carcinoma cell line. Helvetica Chimica Acta 72, 464–471.
Decosterd, L.A., Hoffmann, E., Kyburz, R., Bray, D., Hostettmann, K.,
1991. A new pholoroglucinol derivative from Hypericum calycinum
with antifungal and in vitro antimalarial activity. Planta Medica 57,
548–551.
Dias, A.C.P., Tomas-Barberan, F.A., Fernandes-Ferreira, M., Ferreres, F.,
1998. Unusual flavonoids produced by callus of Hypericum perforatum.
Phytochemistry 48, 1165–1168.
Diplock, A.T., 1997. Will the good fairies please prove us that vitamin E
lessens human degenerative disease? Free Radical Research 27, 511–
532.
Flausino, O.A., Zangrossi, H., Salgado, J.V., Viana, M.B., 2002. Effects of
acute and chronic treatment with Hypericum perforatum L. (LI 160) on
different anxiety-related responses in rats. Pharmacology Biochemistry
and Behaviour 71, 251–257.
Gunatilaka, A.A.L., Balasubramaniam, S., Kumar, V., 1979. 2,3-Di-
metoxyxanthone from Hypericum mysorense. Phytochemistry 18, 182–
183.
Heinoman, M., Rein, D., Satue-Gracia, M.T., Huang, S.W., German, J.B.,
Frankel, E.N., 1998. Effect of protein on the antioxidant activity of
phenolic compounds in a lecithin-liposome oxidation system. Journal
of Agricultural and Food Chemistry 46, 917–922.
Herrera, R.M., Perez, M., MartinHerrera, D.A., LopezGarcia, R., Ra-
banal, R.M., Arias, A., 1996. Antimicrobial activity of extracts from
plants endemic to the Canary Islands. Phytotherapy Research 10, 364–
366.
Hu, L.H., Sim, K.Y., 2000. Sampsoniones A–M, a unique family of
caged polyprenylated benzoylphloroglucinol derivatives, from Hyper-
icum sampsonii. Tetrahedron 56, 1379–1386.
Huang, S.W., Hopia, A., Schwarz, K., Frankel, E.N., German, J.B., 1996.
Antioxidant activity of ␣-tocopherol and trolox in different lipid sub-
strates: bulk oils vs oil-in-water emulsions. Journal of Agricultural and
Food Chemistry 44, 444–452.
Hubner, W.D., Kirste, T., 2001. Experience with St. John’s wort (Hy-
pericum perforatum) in children under 12 years with symptoms of
depression and psychovegetative disturbances. Phytotherapy Research
15, 367–370.
Ishiguro, K., Yamaki, M., Kashihara, M., Takagi, S., Isoi, K., 1990.
Srothralin G: a new antimicrobial compound from Hypericum japon-
icum. Planta Medica 56, 274–276.
Ishiguro, K., Nagata, S., Fukumoto, H., Yamaki, M., Takagi, S., Isoi, K.,
1991. A flavanonol rhamnoside from Hypericum japonicum. Phyto-
chemistry 30, 3152–3153.
Ishiguro, K., Nagata, S., Fukumoto, H., Yamaki, M., Isoi, K., Oyama,
Y., 1993a. An isopentenylated flavonol from Hypericum japonicum.
Phytochemistry 32, 1583–1585.
Ishiguro, K., Yamaki, M., Kashihara, M., Takagi, S., Isoi, K., 1993b. A
chromene from Hypericum japonicum. Phytochemistry 29, 1010–1011.
Ishiguro, K., Nagata, S., Fukumoto, H., Yamaki, M., Isoi, K., 1994.
Pholoroglucinol derivatives from Hypericum japonicum. Phytochem-
istry 35, 469–471.
Ishiguro, K., Nagareya, N., Fukumoto, H., 1998. A phloroglucinol deriva-
tive from cell suspension cultures of Hypericum patulum. Phytochem-
istry 47, 1041–1043.
Ito, N., Fukushima, S., Hassegawa, A., Shibata, M., Ogiso, T., 1983.
Carcinogenicity of butylated hydroxyanisole in F344 rats. Journal of
The National Cancer Institute 70, 343–347.
Jayasuriya, H., McChesney, J.D., Swanson, S.M., Pezzuto, J.M., 1989.
Antimicrobial and cytotoxic activity of rottlerin-type compounds
from Hypericum drummondi. Journal of Natural Products 52, 325–
331.
Johnson, L.F., Jankowski, W.C., 1978. Carbon-13 NMR Spectra—A Col-
lection of Assigned, Coded and Indexed Spectra, Spectrum No. 461.
R.E. Krieger, Huntington, NY.
Kitanov, G.M., 2001. Hypericin and pseudohypericine in some Hypericum
species. Biochemical Systematics and Ecology 29, 171–178.
Lu, C.M., Lin, C.N., 1994. Flavonoids and 9-hydroxytridecyl docosanoate
from Artocarpus heterophyllus. Phytochemistry 35, 781–783.
Maisenbacher, P., Kovar, A., 1992. Adhyperforin: a homologue of
hyperforin from Hypericum perforatum. Planta Medica 58, 291–
293.
Meruelo, D., Lavie, G., Lavie, D., 1988. Therapeutic agents with dramatic
antiretroviral activity and little toxicity at effective doses: aromatic
polycyclic diones hypericin and pseudohypericin. Proceedings of The
National Academy of Sciences of The United States of America 85,
5230–5234.
Rath, G., Potterat, O., Mavi, S., Hostettmann, K., 1996. Xanthones from
Hypericum roeperanum. Phytochemistry 43, 513–520.
Rubenstein, I., Goad, L.J., Clague, A.D.H., Mulheirn, L.J., 1976. The 220
MHz spectra of phytosterols. Phytochemistry 15, 195–200.
Sakar, M.K., Tamer, A.U., 1990. Antimicrobial activity of different extracts
from Hypericum species. Fitoterapia 61, 464–466.
Sanchez-Mateo, C.C., Prado, B., Rabanal, R.M., 2002. Antidepresant
effects of the methanol extract of several Hypericum species from the
Canary Islands. Journal of Ethnopharmacology 79, 119–127.
Shukla, Y.N., Thakur, R.S., 1990. A triterpenoid saponin from Panax
pseudo-ginseng subsp. himalaicus var. angustifolius. Phytochemistry
29, 239–241.
Singleton, V.L., Orthofer, R., Lamuela-Raventos, R.M., 1999. Analysis
of total phenols and other oxidation substrates and antioxidants by
means of Folin Ciocalteu Reagent. Methods in Enzymology 299, 152–
178.
Strolin-Benedetti, M., Dosterd, P.L., 1992. In: B. Testa (Eds.), Ad-
vances in Drug Research, vol. 23. Academic Press, London, pp. 65–
125.
Tada, M., Chiba, K., Yamada, H., Maruyama, H., 1991. Phlorogluci-
nol derivatives as competitive inhibitors against thromboxane A2 and
leukotriene D4 from Hypericum erectum. Phytochemistry 30, 2559–
2562.
Trifunovic, S., Vajs, V., Macura, S., Juranic, N., Djarmati, Z., Jankov,
R., Milosavljevic, S., 1998. Oxidation products of hyperforin from
Hypericum perforatum. Phytochemistry 49, 1305–1310.
Weng, X.C., Wang, W., 2000. Antioxidant activity of compounds isolated
from Salvia plebeia. Food Chemistry 71, 489–493.
 A. Cakir et al. / Journal of Ethnopharmacology 87 (2003) 73–83
Wu, Q.L., Wang, S.P., Du, L.J., Zhang, S.M., Yang, J.S., Xiao, P.G.,
1998a. Chromone glycosides and flavonoids from Hypericum japon-
icum. Phytochemistry 49, 1417–1420.
Wu, Q.L., Wang, S.P., Du, L.J., Yang, J.S., Xiao, P.G., 1998b. Xan-
thones from Hypericum japonicum and H. henryi. Phytochemistry 49,
1395–1402.
Vantieghem, A., Xu, Y., Vandenabeele, P., Denecker, G., Vandenheede,
J.R., Merlevede, W., de Witte, P.A., Agostinis, P., 2001. Different
pathways mediate cytochrom c release after photodynamic therapy
with hypericin. Photochemistry and Photobiology 74, 133–142.
83
Yazaki, K., Okada, T., 1994. Hypericum erectum Thunb. (St. John’s
wort): in vitro culture and the production of procynadins. Biotecnology
in Agriculture and Forestry 26, 167–178.
Yen, G.H., Chen, H.Y., 1995. Antioxidant activity of various tea extracts
in relation to their antimutagenicity. Journal of Agricultural and Food
Chemistry 43, 27–32.
Yildirim, A., Mavi, A., Kara, A.A., 2001. Determination of antioxidant
and antimicrobial activities of Rumex crispus L. extracts. Journal of
Agricultural and Food Chemistry 49, 4083–4089.