Pro-oxidant and cytotoxic activities of atractylenolide I in human promyeloleukemic HL-60 cells

Pro-oxidant and cytotoxic activities of atractylenolide I in

human promyeloleukemic HL-60 cells

Ching-Chiung Wang

, Shyr-Yi Lin

, Huey-Chuan Cheng

, Wen-Chi Hou

b_{Department of Internal Medicine, School of Medicine, Taipei Medical University, Taipei, Taiwan} c

Department of Internal Medicine, Taipei Medical University Hospital, Taipei, Taiwan

d

Mackay Memorial Hospital, Taipei 104, Taiwan

e

Mackay Medicine, Nursing and Management College, Taipei 112, Taiwan Received 12 October 2005; accepted 20 February 2006

Abstract

The dried rhizome of Bai Zhu (Atractylodes ovata) is widely used as a Chinese herbal medicine. Two sesquiterpenolides of similar

structures (atractylenolide I, AT-I; atractylenolide III, AT-III) were isolated from dried rhizome of Atractylodes ovata. Incubation of

AT-I with recombinant human Cu,Zn-superoxide dismutase (rhCu,Zn-SOD) resulted in rhCu,Zn-SOD fragmentations and Zn releases.

However, these were not observed in the AT-III reaction. The AT-1 showed dose-dependent cytotoxic activities (7.5, 15, and 30 lg/ml)

on the human promyeloleukemic HL-60 cells while AT-III did not, and the IC

of the former being 10.6 lg/ml (corresponding to

46 lM) on 12 h-treated cells. The results of DNA ladder and DNA contents in sub-G1 type revealed that AT-I induced apoptosis in

human promyeloleukemic HL-60 cells. The cytotoxic and pharmacological mechanisms of AT-I against human promyeloleukemic

HL-60 cells was investigated. The AT-I appeared to exhibit both pro-oxidant and antioxidant properties after an ESR spectrometer

was used to detect hydroxyl radical productions in vitro and flow cytometry to detect intracellular ROS productions in AT-I treated

cells. The AT-1 also showed dose-dependent Cu,Zn-SOD inhibitory activity in HL-60 cells treated for 12 h, confirmed by activity

and immune stainings. However, catalase, Mn-SOD, and glutathione peroxidase did not apparently change activities under the same

treatments. The addition of commercial rhCu,Zn-SOD (25–100 U/mL) to the AT-I-treated HL-60 cells (15 lg/ml) resulted in significant

differences (p < 0.01) and could reduce the AT-I cytotoxicity from 78% to 28% on HL-60 cells. It was proposed that the AT-I might work

via Cu,Zn-SOD inhibition in HL-60 cells to induce apoptosis and bring about cytotoxicity.

 2006 Elsevier Ltd. All rights reserved.

Keywords: Apoptosis; Atractylodes ovata; Atractylenolide I; Cu,Zn-superoxide dismutase; Flow cytometry; Human promyeloleukemic HL-60 cells; Sesquiterpenoides

1. Introduction

Bai Zhu is the dried rhizome of Atractylodes ovata

(Compositae), a popular traditional Chinese herb used as

a tonic for spleen and stomach ailments in Asia. The

extracts of the herb have been reported to have several

pharmacological activities, such as anti-inflammatory

(

Endo et al., 1979

) and anti-ulcer properties (

Matsuda

et al., 1991; Nogami et al., 1986; Kubo et al., 1983

),

lipid peroxidation inhibition (

Kiso et al., 1983; Kiso

et al., 1985

), and inhibitory activities against tert-butyl

doi:10.1016/j.fct.2006.02.008

Abbreviations: AT-I, atractylenolide I; rhCu,Zn-SOD, recombinant human Cu,Zn-superoxide dismutase; DCFH/DA, dichlorodihydrofluores-cein diacetate; DMPO, 5,5-dimethyl-1-pyrroline-N-oxide; ESR, electron spin-resonance; PAGE, polyacrylamide gel electrophoresis; ROS, reactive oxygen species; Zincon, (2-carboxy-20_-hydroxy-50_{-sulfoformazylbenzene).}

*

Corresponding author. Tel.: +886 2736 1661x6160; fax: +886 (2) 2378 0134.

E-mail address:[email protected](W.-C. Hou).

hydroperoxide-induced cytotoxicity in primary culture of

rat hepatocytes (

Satoh et al., 1996; Bakurai et al., 1993

).

Copper-zinc superoxide dismutase (Cu,Zn-SOD) is a

first-line cytosolic enzyme for protecting cells from

super-oxide radical injury (

Valentine et al., 2005

), and a

supple-ment of Cu,Zn-SOD could increase neuroprotective

effects against ischemic neuronal damage in the gerbil

hip-pocampus (

Hwang et al., 2005

). However, hydrogen

perox-ide (

Choi et al., 1999

) or peroxynitrite (

Alvarez et al., 2004

)

were reported to inactivate Cu,Zn-SOD, and nitric oxide

(

Niketic´ et al., 1999

) was also reported to inactivate

Mn-SOD and Fe-Mn-SOD, and fragmented Mn-SOD was found in

PAGE gels.

Huang et al. (2000)

pointed out that

2-meth-oxyoestradiol and its structural derivatives could selectively

kill human leukemia cells through inhibitions of

Cu,Zn-SOD, and the inhibition of SOD caused accumulation of

cellular superoxide radicals, finally leading to apoptosis.

In a previous study, four structure-related

sesquiter-penes (atractylon and atractylenolides (AT)-I, AT-II and

AT-III) were isolated from A. ovata. Both atractylon and

AT-I showed dose-dependent cytotoxicities against

HL-60 and P-388 cell lines, and the atractylon-treated HL-HL-60

cells were further investigated in the apoptotic parameters

of DNA ladder, sub-G

DNA contents, and PARP

cleav-ages (

Wang et al., 2002

). In this report, the cytotoxic

mech-anism of AT-I against human promyeloleukemic HL-60

cells was investigated in comparison with structure-related

AT-III. We proposed the cytotoxicity of AT-I against

human promyeloleukemic HL-60 cells relative to its

pro-oxidant activity and the inhibition against

Cu,Zn-SOD activity.

2. Materials and methods

2.1. Isolation of AT-I and AT-III from A. ovata

The isolation and structure identification of AT-I and AT-III (Fig. 1A) from A. ovata were according the previous report (Wang et al., 2002). The molecular mass of AT-I and AT-III was 230.1 Da and 248.1 Da, respec-tively. The difference between AT-I and AT-III was that the former had a double bond between the C-8 and C-9 positions while the latter had a hydroxyl group in the C-8 position.

2.2. Cell cultures

Human promyeloleukemic HL-60 cells were obtained from American Type Cell Culture (ATCC) (Rockville, MD, USA). The human promy-eloleukemic HL-60 cells were maintained in RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and 100 mg/l streptomycin and were incubated at 37C in a humidified atmosphere of 5% CO2.

2.3. Cell viability

A stock solution of AT-I and AT-III (20 mg/ml) was prepared by dissolving them in DMSO and then storing at20 C until use. Serial dilutions of these stock solutions were prepared in the culture medium in 24-well microtiter plates. AT-I or AT-III at different concentrations (7.5, 15, and 30 lg/ml, respectively corresponding to 32.6, 65.2, and 130.4 lM for AT-I; and 30.2, 60.5, and 120.9 lM for AT-III) was added to cell cultures for 12 h without renewal of the medium. The cell viability was assayed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide

(MTT) staining (Allen et al., 1997). To prevent AT-I toxicity (15 lg/mL) to HL-60 cells, the rhCu,Zn-SOD (25, 50, and 100 U/ml, 574595, Cal-biochem Co.) was pre-treated for 1 h and then co-cultured with AT-I for 12 h and assayed for cell viability.

2.4. DNA isolation and electrophoresis

The human promyeloleukemic HL-60 cells (1· 106

cells/well) were treated with AT-I at different concentrations (7.5, 15, and 30 lg/ml) for 12 h. Cells were collected, washed with PBS twice and then lysed in 100 ml of lytic buffer (50 mM Tris, pH 8.0; 10 mM EDTA; 0.5% sodium sarko-sinate, and 1 mg/mL proteinase K) for 3 h at 56C. DNA was then extracted with phenol/chloroform/isoamyl alcohol (V/V/V, 25/24/1) before loading. The extent of DNA fragmentation was assessed by 1.5% agarose gel electrophoresis (Allen et al., 1997).

2.5. Flow cytometry analysis of sub-G

DNA contents

The AT-I-treated (7.5, 15, and 30 lg/ml) HL-60 cells (5· 105_cells/

well) were harvested by centrifugation and washed with PBS. They were then fixed with ice-cold 80% ethanol for at least 30 min and then stained with propidium iodide (Sigma). Total fluorescence intensities were quan-tified by a FACScan flow cytometer (Becton Dickinson, CA, USA). The results of sub-G1DNA contents were calculated as the number of cells

versus the amount of DNA as indicated by the intensity of fluorescence (Allen et al., 1997).

2.6. Measurement of intracellular peroxide levels by

flow cytometry

The intracellular peroxide levels were assessed by flow cytometry using dichlorodihydrofluorescein diacetate (DCFH/DA, Molecular Probes) as a probe (Gorman et al., 1997). The DCFH/DA penetrated into cells and was hydrolyzed by cellular esterase to DCFH, which was further oxidized by intracellular peroxide into a strong fluorescent compound, dichloro-fluorescein. The AT-I (0.125, 0.25, and 0.5 lg/ml), or 0.015% H2O2

(positive controls) were used to treat human promyeloleukemic HL-60 cells (1· 105_{cells/well) in the culture medium for 30 min, and then cells}

were cultured for another 30 min after the addition of 5 lM DCFH/DA (15 mM in DMSO as a stock solution). The cells were washed twice with PBS and quantified by a FACScan flow cytometer (Becton Dickinson, CA, USA) with excitation and emission settings of 488 and 530 nm, respectively. The peroxide levels in cells were plotted as one-parameter histograms with cell count on the y-axis and the fluorescence on the x-axis. The calculated area of fluorescent distributions in the fixed cell counts were expressed as the fluorescent intensity.

2.7. Western blots and activity staining

Crude extracts (30 lg proteins) of AT-I-treated or AT-III-treated HL-60 cells were harvested and separated on 10% native PAGE gels and then were stained either for catalase (Woodbury et al., 1971) or SOD (Beau-champ and Fridovich, 1971) activity or transferred onto Hybond-P PVDF membranes (Amersham Pharmacia Biotech Co.) for Cu,Zn-SOD, Mn-SOD, or glutathione peroxidase immune staining, which was detected by anti-Cu/Zn SOD (574597, Calbiochem Co.), anti-Mn SOD (574596, Calbiochem Co.), or anti-glutathione peroxidase (CR2144SP, Cortex Biochem., CA, USA) antibodies, respectively.

2.8. SOD inhibition and released Zn determination

Each of the 30 and 60 lg of AT-I or AT-III were pre-mixed with 10 U of commercial recombinant human Cu,Zn-SOD (574595, Calbiochem Co.) for 12 h and were separated on 10% native PAGE gels. The gel was stained for SOD activity (clear bands against purple backgrounds, (Beauchamp and Fridovich, 1971)) or transferred onto Hybond-P PVDF

membranes for Cu,Zn-SOD immune staining to check the SOD inhibition and fragmentation. For released Zn determinations, the 50 lg AT-I was pre-mixed with 50, 100, and 150 U of rhCu,Zn-SOD in the total 35 ll mixture for 12 h, and the released Zn from rhCu,Zn-SOD was determined by Zincon reagent (2-carboxy-20_-hydroxy-50_{-sulfoformazylbenzene)}

(Pattanaik et al., 1992) to form Zn–Zincon complex. The absorbance at 620 nm was determined, and the only rhCu,Zn-SOD was used as a blank. The ZnCl2was used to plot the standard curve.

2.9. Pro-oxidant and antioxidant properties of AT-1 in Fenton

reaction detected by ESR spectrometer

The hydroxyl radical was generated by Fenton reaction according to the method of Kohno et al. (1991). The total 500 lL mixture included 0.02, 0.025, 0.0375, and 0.05 lg/ml of AT-I, 5 mM 5,5-dimethyl-1-pyrro-line-N-oxide (DMPO) and 0.05 mM ferrous sulfate. After mixing, the solution was transferred to an ESR quartz cell and placed at the cavity of the ESR spectrometer, and then hydrogen peroxide was added to a final concentration of 0.25 mM. Deionized water was used instead of sample solution for blank experiments. After forty seconds, the intensity of the signal of DMPO-OH spin-adduct (ISADMPO-OH) was measured. All ESR

spectra were recorded at the ambient temperature (298 K) on a Bruker EMX-6/1 ESR spectrometer equipped with WIN-ESR SimFonia soft-ware, Version 1.2. The conditions of ESR spectrometry were as follows: center field, 345.4 ± 5.0 mT; microwave power, 8 mW (9.416 GHz); modulation amplitude, 5 G; modulation frequency, 100 kHz; time

con-stant, 0.6 s; scan time, 1.5 min. The intensities of DMPO-OH spin signal in ESR spectrometry were used to evaluate the pro-oxidant or antioxidant properties of AT-I and were calculated as the peak height which was standardized by WIN-ESR SimFonia software. The calculated DMPO-OH adducts in the control was assumed as 100%.

2.10. Statistics

Means of triplicates were measured. Student’s t-test was used for comparison between two treatments. A difference between the control and each treatment was considered statistically significant when P < 0.05 (*_{) or}

P < 0.01 (**_).

3. Results

3.1. Induction of apoptosis of HL-60 cells by AT-1

Though the difference between AT-I and AT-III

struc-tures is minor (

Fig. 1

A), the AT-1 showed dose-dependent

cytotoxic activities (7.5, 15, and 30 lg/ml,

Fig. 1

B) on the

human promyeloleukemic HL-60 cells while AT-III did

not, and the IC

of the former was 10.6 lg/ml

(corre-sponding to the 46 lM) on 12 h-treated cells. The apoptotic

parameters, such as DNA ladders (

Fig. 1

C) and sub-G

Fig. 1. The cytotoxic activity of AT-I and AT-III on the HL-60 cells. (A) The structures of AT-I and AT-III, (B) the cytotoxicity index of AT-I and AT-III in the concentrations of 7.5, 15, and 30 lg/ml on 12 h-cultured HL-60 cells, (C) the apoptotic parameter of DNA ladder of AT-I-treated HL-60 cells for 12 h in the concentrations of 7.5, 15, and 30 lg/ml, (D) the apoptotic parameter of sub-G1DNA contents (M1) of AT-I-treated HL-60 cells for 12 h in the

DNA contents (

Fig. 1

D), showed that AT-1 induced

apop-tosis and resulted in cytotoxicity to human

promyeloleuke-mic HL-60 cells. The values most below the G

peak in the

DNA contents (M1) were 4.7%, 13.5%, 21.7%, and 42.5%,

respectively, for controls, 7.5, 15, and 30 lg/ml of

AT-I-treated HL-60 cells.

3.2. SOD inhibition and released Zn determination

Recombinant human Cu,Zn-SOD was inhibited by

AT-1, as observed by activity staining (faint clear bands,

Fig. 2

A) and immune staining (

Fig. 2

B). Each of the

30 lg (lanes 1–3) and 60 lg (lanes 2–4) of AT-I (lanes 1

and 2) or AT-III (lanes 3 and 4) were pre-mixed with

10 U of commercial rhCu,Zn-SOD for 12 h and were

sepa-rated on 10% native PAGE gels. Compared to the control

(Cu,Zn-SOD only, lane C), it was found that AT-1, but not

AT-III, could inactivate rhCu,Zn-SOD (faint clear bands,

lanes 1 and 2,

Fig. 2

A) and resulted in SOD fragmented

patterns (lanes 1 and 2,

Fig. 2

A and B). The 60 lg of

AT-I (lane 2,

Fig. 2

A and B) made more serious SOD

frag-mentations than 30 lg did (lane 1,

Fig. 2

A and B). The

released Zn contents from Cu,Zn-SOD were determined

using a Zn–Zincon complex standard curve (

Fig. 2

C). It

was found that the released Zn contents increased

(

Fig. 2

D) with increased rhCu,Zn-SOD (50, 100, and

150 U) at a fixed amount of AT-I (50 lg). The released

Zn from rhCu,Zn-SOD was 3.92, 5.86, and 15.59 lg/ml,

respectively, for 50, 100, and 150 U of rhCu, Zn-SOD.

3.3. The changes of antioxidant enzymes in AT-I-treated

or AT-III-treated HL-60 cells

The human promyeloleukemic HL-60 cells were treated

with AT-I (lanes 1–3 were 7.5, 15, and 30 lg/ml,

respec-tively) or AT-III (lanes 4–6 were 7.5, 15, and 30 lg/ml,

respectively) for 12 h. Each cell extract (30 lg protein)

was separated on 10% native PAGE gels for Cu,Zn-SOD

activity staining (clear band against the purple

back-grounds,

Fig. 3

A) or immune stainings of Cu,Zn-SOD

(

Fig. 3

B) and Mn-SOD (

Fig. 3

C). Compared to the control

lane (lane C, without AT-I or AT-III treatments),

Cu,Zn-SOD dose-dependent inhibitions were found in

AT-I-trea-ted cells (lanes 1–3,

Fig. 3

A, activity staining; lanes 1–3,

Zn concentration (μg/mL) 0 5 10 15 20 25 30 A 620 nm (Zn-Zincon complex) 0.0 0.1 0.2 0.3 0.4 C D Released Zn contents (μg/mL) 0 3 6 9 12 15 18

AT-I treated rhCu, Zn-SOD (unit)

0 20 40 60 80 100 120 140 160

Fig. 2. The effects of AT-I on recombinant human Cu,Zn-SOD activity detected by (A) activity staining (clear bands against purple background) or (B) immune staining. The 30 lg (lanes 1–3) and 60 lg (lanes 2–4) of AT-I (lanes 1 and 2) or AT-III (lanes 3 and 4) were pre-mixed with 10 U of commercial rhCu,Zn-SOD for 12 h and were separated on 10% native PAGE gels. (C) The ZnCl2was used to plot the standard curve as detected by Zincon reagent to

form Zn–Zincon complex (absorbance at 620 nm). (D) The released Zn from rhCu,Zn-SOD (50, 100 and 150 U) after AT-I treatments was calculated from the Zn standard curve, and the only rhCu,Zn-SOD was used as a blank.

Fig. 3

B, immune staining). AT-III, however, did not show

any apparent inhibitory activities on Cu,Zn-SOD. Neither

AT-I nor AT-III treatments inhibited the expression of

Mn-SOD activity in HL-60 cells (

Fig. 3

C, immune

stain-ing). AT-I treatments also did not have any apparent

impact on glutathione peroxidase (lanes 1–3,

Fig. 3

D,

immune staining) or catalase (lanes 1–3,

Fig. 3

E, activity

staining, yellow band against the deep blue backgrounds

)

activities.

3.4. Pro-oxidant and antioxidant properties of AT-1 in

Fenton reaction detected by ESR spectrometer

The intensities of DMPO-OH adducts in the magnetic

field from 3426 to 3526 Gauss were used to detect the

hydroxyl radicals generated by the Fenton reaction. The

intensities of DMPO-OH spin signal in ESR spectrometry

were used to evaluate the pro-oxidant or antioxidant

prop-erties of AT-I and were calculated as the peak height which

was standardized by WIN-ESR SimFonia software. The

calculated DMPO-OH adducts in the control was assumed

as 100%. AT-I was added in the Fenton reaction system to

evaluate the promotion (pro-oxidant) or inhibition

(antiox-idant) properties on hydroxyl radical generations in vitro.

The relative intensity of calculated DMPO-OH adducts

(

Fig. 4

) in the control was assumed to be 100%, and the

added AT-I of 0.02 (B), 0.025 (C), 0.0375 (D), and

0.05 lg/ml

(E)

showed

83.92%

(p < 0.01),

102.07%,

124.73% (p < 0.01), and 55.68% (p < 0.01) relative

intensi-ties, respectively. The AT-I showed pro-oxidant effects on

hydroxyl radical productions in the concentrations of

0.025–0.0375 lg/ml, and antioxidant effects in the

concen-trations higher than 0.05 lg/ml. From the intensity data

in the ESR spectra it can be seen that the AT-I exhibited

pro-oxidant and antioxidant properties on hydroxyl radical

generations in the Fenton reaction system.

3.5. Measurement of intracellular peroxide levels in

AT-I treated HL-60 cells

The DCF fluorescent intensity was subjected to flow

cytometric analysis to assess the intracellular peroxide

levels in each set of AT-I treated HL-60 cells. The peroxide

levels in cells were plotted as one-parameter histograms

with cell count on the y-axis and the fluorescence on the

x-axis. The calculated area of fluorescent distributions in

the fixed cell counts were expressed as the fluorescent

inten-sity (

Fig. 5

). Peroxide levels in the untreated control was

187.69 (A), and the fluorescent intensities of AT-I-treated

cells in the concentrations of 0.125 (B), 0.25 (C), and

0.5 lg/ml (D) were 410.47 (p < 0.01), 319.08 (p < 0.01),

and 224.68 (p < 0.05), respectively. All of the tested

concen-trations increased the peroxide levels in HL-60 cells.

Hydro-gen peroxide (E, positive control, 0.015%) dramatically

AT-III-treated HL-60 cells for 12 h. The activity stainings (clear bands against the dark background) of Cu,Zn-SOD (A) and catalase (E) or immune stainings of Cu,Zn-SOD (B), Mn-SOD (C), and glutathione peroxidase (D). The HL-60 cells were treated with AT-I (lanes 1–3 were 7.5, 15, and 30 lg/ml, respectively) or AT-III (lanes 4–6 were 7.5, 15, and 30 lg/ml, respectively) for 12 h. Each cell extracts (30 lg protein) was separated on 10% native PAGE gels for activity stainings or transfer onto PVDF membrane for immune stainings.

AT-I (μg/mL) Intensity of DMPO-OH adducts (%) 0 10 20 30 40 50 60 90 100 110 120 130 140 Control 0.02 0.025 0.0375 0.05 (A) (B) (C) (D) (E)

**

**

**

Fig. 4. Pro-oxidant and antioxidant properties of AT-1 in Fenton reaction detected by ESR spectrometer in the magnetic field from 3426 to 3526 Gauss. The intensity of DMPO-OH adduct was measured without (as the control, A) or with AT-I additions in the concentrations of 0.02 (B), 0.025 (C), 0.0375 (D), and 0.05 lg/ml (E). The intensities of DMPO-OH spin signal in ESR spectrometry were used to evaluate the pro-oxidant or antioxidant properties of AT-I and were calculated as the peak height which was standardized by WIN-ESR SimFonia software. The calculated DMPO-OH adducts in the control was assumed as 100%, the relative intensities were 83.92%, 102.07%, 124.73%, and 55.68%, respectively, for 0.02, 0.025, 0.0375, and 0.05 lg/ml. A difference between the control and each treatment was considered statistically significant when p < 0.05 (*_{) or}

p < 0.01 (**_).

1 _{For interpretation of color in this figure, the reader is referred to the}

increased the intracellular fluorescent levels. We found the

AT-I treatments to exhibit reverse effects on the

intracellu-lar peroxide levels, with the lower concentration (0.125 lg/

ml) of AT-I increasing them. However, the higher

concen-tration of AT-I (0.5 lg/ml) decreased them in line with the

untreated ones.

3.6. Prevention of cytotoxicity of AT-I-treated HL-60

cells by adding Cu,Zn-SOD

The AT-I (15 lg/ml) treatment resulted in the death of

78% of the human promyeloleukemic HL-60 cells.

How-ever, the pre-treatment of rhCu,Zn-SOD (25, 50, 100 U/

ml) for 1 h had a significant impact (p < 0.01) and

decreased the cytotoxicity to 28% (100 U/ml) (

Fig. 6

).

4. Discussion

Bai Zhu is the dried rhizome of A. ovata (Compositae), a

popular traditional Chinese herb used as a tonic for spleen

and stomach ailments in Asia. Although several

pharmaco-logical activities were reported (

Endo et al., 1979; Matsuda

et al., 1991; Nogami et al., 1986; Kubo et al., 1983;

Kiso et al., 1983; Kiso et al., 1985; Satoh et al., 1996;

Bak-urai et al., 1993

), however, few reports were concerned for

anticancer treatments. In the previous report (

Wang et al.,

2002

), the sesquiterpenes (atractylenolides (AT)-I, and

AT-III) were isolated from A. ovata. The molecular mass of

AT-I and AT-III was 230.1 Da and 248.1 Da, respectively.

The difference between AT-I and AT-III was that the

for-mer had a double bond between the C-8 and C-9 positions

while the latter had a hydroxyl group in the C-8 position.

The AT-I was reported to have anticancer activity (

Wang

et al., 2002

), therefore, in the present study, the cytotoxic

mechanism of AT-I against human promyeloleukemic

HL-60 cells was investigated. We proposed the AT-I

against HL-60 cells relative to pro-oxidant and inhibition

against Cu,Zn-SOD activity.

Though AT-I and AT-III had similar structures, the

for-mer exhibited apparent cytotoxicity to HL-60 cells under

the same concentrations. AT-I showed dose-dependent

cytotoxicities to HL-60 cells (IC

, 10.6 lg/ml

correspond-ing 46 lM for 12 h-treated cells), and the DNA ladder and

increased sub-G

DNA contents revealed that AT-I

induced apoptosis (

Fig. 1

). The IC

(46 lM) of AT-I in

cells treated for 12 h was lower than that of myricetin,

apigenin, and close to that of baicalein and fisetin (

Lee

et al., 2002

). The IC

of taraxinic acid (a hydrolysate

ses-quiterpene lactone glycoside) was 34.5 lM for the 48-h

treated HL-60 (

Choi et al., 2002

) which was toxic than that

of the AT-I, however, the treatment time (48 h) was longer

than the present report (12 h). Several sesquiterpenoids

iso-lated from Pulicaria canariensis had IC

values higher

than that of AT-I in HL-60 cells treated for 12 h (

Triana

et al., 2005

). Citrinin had cytotoxic activity toward

HL-60 cells, and the IC

value was close to 50 lM in 24

h-trea-ted cells (

Yu et al., 2006

). From above data, AT-I had

potent cytotoxicity toward HL-60 cells. Investigations into

the AT-I-induced cytotoxic mechanism and into the effects

of AT-I on antioxidant enzymes, on ROS productions in

the treated cell system, and on hydroxyl radical

produc-tions in the cell-free system continue.

Control _0.125 0.25 0.5 H2O2 AT-I (μg/mL)

**

**

**

*

Fluorescent intensity

Fig. 5. Measurement of intracellular peroxide levels in AT-I treated HL-60 cells by flow cytometric analysis using DCFH/DA fluorescent probes. The intracellular peroxide levels without (A) or with AT-I treated HL-60 cells in the concentrations of 0.5 (B), 0.25 (C), and 0.125 lg/ml (D) for 30 min, and the 0.015% of hydrogen peroxide (E) was used as a positive control. The peroxide levels in cells were transferred as one-parameter histograms with cell count on the y-axis and the fluorescence on the x-axis. The calculated area of fluorescent distributions in the fixed cell counts were expressed as the fluorescent intensity. A difference between the control and each treatment was considered statistically significant when p < 0.05 (*_{) or p < 0.01 (}**_).

Fig. 6. The effects of added rhCu,Zn-SOD on cytotoxicity of AT-I-treated HL-60 cells. Pre-treatment of rhCu,Zn-SOD (25, 50, 100 U/ml) 1 h before AT-I-treated HL-60 cells (15 lg/ml) decrease the cytotoxicity from 78% to 28% (100 U/ml). A difference between the control and each treatment was considered statistically significant when p < 0.05 (*_{) or p < 0.01 (}**_).

Cu,Zn-SOD is a first-line cytosolic enzyme for

protect-ing cells from superoxide radical injury (

Valentine et al.,

2005

), and the Cu,Zn-SOD supplements could increase

neuroprotective effects against ischemic neuronal damage

in the gerbil hippocampus (

Hwang et al., 2005

). However,

hydrogen peroxide (

Choi et al., 1999

) or peroxynitrite

(

Alvarez et al., 2004

) were reported to inactivate

Cu,Zn-SOD and fragmented Cu,Zn-SOD was found in PAGE gels. Choi

et al. proposed that hydroxyl radicals resulted in SOD

frag-mentations and could be recovered by carnosine,

homocar-nosine, and anserine (

Choi et al., 1999

). Therefore, different

amounts of AT-I were pre-mixed with rhCu, Zn-SOD and

separated on native PAGE gels. It was found that AT-I

inhibited rhCu,Zn-SOD activity and resulted in SOD

frag-mentations detected by activity (faint clear bands) and

immune stainings (

Fig. 2

A and B). The released Zn

increased with more rhCu, Zn-SOD (50–150 U) added at

the fixed AT-I concentration (

Fig. 2

D).

Pattanaik et al.

(1992)

used the dichlorodiammineplatinum (II) to react

with Zn-metallothionein, and the released Zn was

deter-mined by the Zincon reagents. The present results

(

Fig. 2

) suggested that the released Zn might be from the

fragmented rhCu, Zn-SOD by AT-I treatments.

In the AT-I-treated cell system, several antioxidant

enzymes—such as Cu,Zn-SOD, Mn-SOD, glutathione

per-oxidase, and catalase (

Fig. 3

)—were detected either by

activity staining or immune staining to allow us to observe

the enzyme changes. It was found that only Cu,Zn-SOD

(

Fig. 3

A, activity stainings;

Fig. 3

B, immune stainings)

was apparently inactivated in cells treated for 12 h.

Huang

et al. (2000)

pointed that 2-methoxyoestradiol and its

struc-tural derivatives could selectively kill human leukemia cells

through inhibitions of Cu,Zn-SOD, and the inhibition of

SOD caused accumulation of cellular superoxide radicals,

finally leading to apoptosis.

Ueda et al. (2001)

reported that baicalin acted as a

pro-oxidant and induced apoptosis of Jurkat cells. The

procy-anidin B2 showed dual antioxidant and pro-oxidant effects

on metal-mediated DNA damage in HL-60 cells (

Sakano

et al., 2005

). The hydroxyl radical productions with or

without AT-I additions in the cell-free Fenton reaction

system detected by ESR spectrometer (

Fig. 4

) or changes

of intracellular peroxide levels in the AT-I-treated HL-60

cell system detected by flow cytometric analysis (

Fig. 5

)

were investigated. Both cell-free (

Fig. 4

) and cell-treated

(

Fig. 5

) systems revealed that AT-I acted as a pro-oxidant

in lower concentrations and as an antioxidant in higher

concentrations, which was similar to the behavior of uric

acid (

Abuja, 1999; Filipe et al., 2002

).

On the other hand, the cytotoxicities of AT-I-treated

cells

could

be

recovered

by

adding

rhCu,Zn-SOD

(

Fig. 6

).

Choi et al. (1999)

proposed that hydroxyl radicals

resulted in SOD fragmentations and could be recovered by

carnosine, homocarnosine, and anserine. In the present

results, AT-I was able to directly inhibit rhCu,Zn-SOD

and cause SOD fragmentations (30 and 60 lg). The AT-I

behaved as an pro-oxidant in the Fenton reaction system

in the ranges of 0.025–0.0375 lg/ml; the AT-I elevated

intracellular peroxide levels after 30 min treatment in the

ranges of 0.125–0.25 lg/ml; the dose-dependent

cytotoxic-ity of AT-I (the actual amounts in cultured medium were

7.5, 15, and 30 lg) in cells treated 12 h accompanying

Cu,Zn-SOD inhibitions. The sesquiterpene-related

com-pounds were hydrophobic properties in nature. If AT-I

can diffuse from medium into cells, then it can inhibit the

Cu,Zn-SOD and result in ROS accumulations and finally

the apoptosis of 60 cell. If it can not diffuse into

HL-60 cells, the pro-oxidant effects in hydroxyl radical

produc-tions of AT-I in the cultured medium may attack the

membranes and elevate the intracellular peroxide levels

(may be via the inhibition of Cu,Zn-SOD), finally resulting

in the apoptosis of HL-60 cell.

Acknowledgement

The authors want to thank the financial support from

the National Science Council, Republic of China (NSC

94-2313-B038-001) and from Committee on Chinese

Med-icine and Pharmacy, Department of Health, Republic of

China (CCMP 93-RD-033).

References

Abuja, P.M., 1999. Ascorbate prevents prooxidant effects of urate in oxidation of human low density lipoprotein. FEES Lett. 446, 305–308. Allen, R.T., Hunter III, W.J., Agrawal, D.K., 1997. Morphological and biochemical characterization and analysis of apoptosis. J. Pharmacol. Toxicol. Meth. 37, 215–228.

Alvarez, B., Demichli, V., Dura´n, R., Trujillo, M., Cerven˜ansky, C., Freeman, B.A., Radi, R., 2004. Inactivation of human Cu,Zn superoxide diamutase by peroxynitrite and formation of histidinyl radical. Free Rad. Biol. Med. 37, 813–822.

Bakurai, T., Yamada, H., Saito, K., Kano, Y., 1993. Enzyme inhibitory activities of acetylene and sesquiterpene compounds in Atractylodes rhizome. Biol. Pharm. Bull. 16, 142–145.

Beauchamp, C., Fridovich, I., 1971. Superoxide dismutase: improved assays and an assay applicable to acrylamide gels. Anal. Biochem. 44, 276–287.

Choi, S.Y., Kwon, H.Y., Kwon, O.B., Kang, J.H., 1999. Hydrogen peroxide-mediated Cu,Zn-superoxide dismutase fragmentation: pro-tection by carnosine, homocarnosine and anserine. Biochem. Biophys. Acta 1472, 651–657.

Choi, J.H., Shin, K.M., Kim, N.Y., Hong, J.P., Lee, Y.S., Kim, H.J., Park, H.J., Lee, K.T., 2002. Taraxinic acid, a hydrolysate of sesquiterpene lactone glycoside from Taraxacum coreanum Nakai, induces the differentiation of human acute promyelocytic leukemia HL-60 cells. Biol. Pharm. Bull. 25, 1446–1450.

Endo, K., Taguchi, T., Taguchi, F., Hikino, H., Yamahara, J., Fujimura, H., 1979. Antiinflammatory principles of Atractylodes rhizomes. Chem. Pharm. Bull. 27, 2954–2958.

Filipe, P., Haigle, J., Freitas, J., Fernandes, A., Mazie`re, J., Mazie`re, C., Santus, R., Morlie`re, P., 2002. Anti- and pro-oxidant effects of urate in copper-induced low-density lipoprotein oxidation. Eur. J. Biochem. 269, 5474–5483.

Gorman, A., McGowan, A., Cotter, T.G., 1997. Role of peroxide and superoxide anion during tumor cell apoptosis. FEES Lett. 404, 27– 33.

Huang, P., Feng, L., Oldham, E.A., Keating, M.J., Plunkett, W., 2000. Superoxide dismutase as a target for the selective killing of cancer cells. Nature 407, 390–395.

Hwang, I.K., Eum, W.S., Yoo, K.Y., Cho, J.H., Kim, D.W., Choi, S.H., Kang, T.C., Kwon, O.S., Kang, J.H., Choi, S.Y., Won, M.H., 2005. Copper chaperone for SOD supplement potentiates the Cu,Zn-SOD function of neuroprotective effects against ischemic neuronal damage in the gerbil hippocampus. Free Rad. Biol. Med. 39, 392–402. Kiso, Y., Tohkin, M., Hikino, H., 1983. Antihepatotoxic principles of

Atractylodes rhizomes. J. Nat. Prod. 46, 651–654.

Kiso, Y., Tohkin, M., Hikino, H., 1985. Mechanism of antihepatotoxic activity of atractylon. I: Effect on free radical generation and lipid peroxidation. Planta Medica 51, 97–100.

Kohno, M., Yamada, M., Mitsuta, K., Mizuta, Y., Yoshikawa, T., 1991. Spin-trapping studies on the reaction of iron complexes with peroxides and the effects of water-soluble antioxidants. Bull. Chem. Soci. Japan 64, 1447–1453.

Kubo, M., Nogami, M., Nishimura, M., Moriura, T., Arichi, S., 1983. Studies of crude drugs about its origin, process and quality. I. The preventive effects of Chinese crude drug ‘‘Zhu’’ on experimental stomach ulcer and its pharmacological evaluation (1). Yakugaku Zasshi 103, 442–448.

Lee, W.R., Shen, S.C., Lin, H.Y., Hou, W.C., Yang, L.L., Chen, Y.C., 2002. Wogonin and fisetin induce apoptosis in human promyeloleu-kemic cells, accompanied by a decrease of reactive oxygen species, and activation of caspase 3 and Ca2+_{-dependent endonuclease. Biochem.}

Pharmacol. 63, 225–236.

Matsuda, H., Li, Y.H., Taniguchi, K., Yamahara, J., Tamai, Y., 1991. Imaging analysis of antiulcer action and the active constituent of atractylodis rhizoma. Yakugaku Zasshi 111, 36–39.

Niketic´, V., Stojanovic´, S., Nikolic´, A., Spasic´, M., Michelson, A.M., 1999. Exposure of Mn and FeSODs, but not Cu/Zn SOD, to NO leads to nitrosonium and nitroxyl ions generation which cause enzyme modification and inactivation: an in vitro study. Free Rad. Biol. Med. 27, 992–996.

Nogami, M., Iwanaga, M., Moriura, T., Kubo, M., 1986. Studies on the origin, processing and quality of crude drugs V pharmacological

evaluation of the Chinese drug ‘‘Zhu’’ in experimental gastric ulcer (5) the preventive effects of Atractylodes ovata on stress-induced gastric ulcer. Yakugaku Zasshi 106, 498–503.

Pattanaik, A., Bachowski, G., Laib, J., Lemkuil, D., Shaw III, C.F., Petering, D.H., Hitchcock, A., Saryan, L., 1992. Properties of the reaction of cis-dichlorodiammineplatinum (II) with metallothionein. J. Biol. Chem. 267, 16121–16128.

Sakano, K., Mizutani, M., Murata, M., Oikawa, S., Hiraku, Y., Kawanishi, S., 2005. Procyanidin B2 has anti- and pro-oxidant effects on metal-mediated DNA damage. Free Rad. Biol. Med. 39, 1041– 1049.

Satoh, K., Nagai, F., Ushiyama, K., Kano, I., 1996. Specific inhibition of Na+, K(+)-ATPase activity by atractylon, a major component of Byaku-Jutsu, by interaction with enzyme in the E2 state. Biochem. Pharmacol. 51, 339–343.

Triana, J., Lo´pez, M., Pe´rez, F.J., Gonza´lez-Platas, J., Quintana, J., Este´vez, F., Leo´n, F., Bermejo, J., 2005. Sesquiterpenoids from Pulicaria canariensis and their cytotoxic activities. J. Nat. Prod. 68, 523–531.

Ueda, S., Nakamura, H., Masutani, H., Sasada, T., Takabayashi, A., Yamaoka, Y., Yodoi, J., 2001. Baicalin induces apoptosis via mitochondrial pathway as prooxidant. Mol. Immunol. 38, 781–791. Valentine, J.S., Doucette, P.A., Potter, S.Z., 2005. Copper-zinc superoxide

dismutase and amyotrophic lateral sclerosis. Ann. Rev. Biochem. 74, 563–593.

Wang, C.C., Chen, L.G., Yang, L.L., 2002. Cytotoxic activity of sesquiterpenoids from Atractylodes ovata on leukemia cell lines. Planta Medica 68, 204–208.

Woodbury, W., Spencer, A.K., Stahmann, M.A., 1971. An improved procedure using ferricyanide for detecting catalase isozymes. Anal. Biochem. 44, 301–305.

Yu, F.Y., Liao, Y.C., Chang, C.H., Liu, B.H., 2006. Citrinin induces apoptosis in HL-60 cells via activation of the mitochondrial pathway. Toxicol. Lett. 161, 143–151.