Molecular Mechanisms of Apoptosis Induced by Magnolol in Colon and Liver Cancer Cells

Molecular Mechanisms of Apoptosis Induced by Magnolol in Colon and Liver Cancer Cells

Shyr-Yi Lin,

Yu-Tza Chang,

Jean-Dean Liu,

Chung-Hsun Yu,

Yuan-Soon Ho,

Yi-Hsuan Lee,

and Wen-Sen Lee

*

1

Graduate Institute of Medical Sciences, Taipei Medical University, Taipei, Taiwan

2

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

3

Graduate Institutes of Cellular and Molecular Biology, Taipei Medical University, Taipei, Taiwan

4

Graduate Institutes of Biomedical Technology, Taipei Medical University, Taipei, Taiwan

5

Department of Physiology, School of Medicine, Taipei Medical University, Taipei, Taiwan

Magnolol has been reported to have anticancer activity. In this study we found that treatment with 100 mm magnolol induced apoptosis in cultured human hepatoma (Hep G2) and colon cancer (COLO 205) cell lines but not in human untransformed gingival fibroblasts and human umbilical vein endothelial cells. Our investigation of apoptosis in Hep G2 cells showed a sequence of associated intracellular events that included (a) increased cytosolic free Ca

; (b) increased translocation of cytochrome c (Cyto c) from mitochondria to cytosol; (c) activation of caspase 3, caspase 8, and caspase 9; and (d) downregulation of bcl-2 protein. Pretreatment of the cells with the phospholipase C inhibitor 1-[6-[[(17b)-3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1 H-pyrrole-2,5-dione (U73122) or the intracellular chelator of Ca

1,2-bis(2-aminophenoxy)ethane-N,N,N

,N

-tetraacetic acid acetoxymethyl ester (BAPTA/AM) inhibited the subsequent magnolol augmentation of [Ca

]

and also the activation of caspase-8 and caspase-9, so that the occurrence of apoptosis in those cells was greatly reduced. Pretreatment of the cells with ZB4 (which disrupts the Fas response mechanism) also decreased the subsequent magnolol-induced caspase-8 activation and reduced the occurrence of apoptosis. We interpreted these findings to indicate that the above-listed sequence of intracellular events led to the apoptosis seen in Hep G2 cells and that [Ca

]

, Cyto c, and Fas function as intracellular signals to coordinate those events.

Key words: caspases; cytochrome c; bcl-2; calcium; Fas

INTRODUCTION

As stated in the recent review by Akriviadis et al.

[1], liver and colorectal cancers remain prevalent, deadly, and increasingly costly to patients and to society. Since there are many complex causative factors in these cancers, the ongoing search for effective therapies most often is based on new discoveries about the underlying cellular mechan- isms. The focus most recently has been on the mechanisms concerned with cell proliferation (mitosis) and the opposing mechanisms leading to programmed cell death (apoptosis). Concurrently, studies also are aimed at discovering pharmaceutical agents that might interfere with these cellular mechanisms in cancerous cells so as to block the growth of tumors and thereby ®nd use as thera- peutic agents.

One such pharmaceutical agent is magnolol, a phenolic compound isolated in pure form from the bark of the tree Magnolia of®cinalis. Magnolol, also known as Hou p'u among Chinese herbalists, has been found to be effective in causing the regression of (or blocking of the formation of) skin papilloma tumors in mice treated with the carcinogens 7,12- dimethylbenz[a]anthracene and 12-O-tetradeca- noylphorbol-13-acetate [2]. Since so little else is

known about the details of magnolol action and the cellular mechanisms involved, we chose to examine this matter in cultures of human cancer cells.

Previously, Wang and Chen [3] reported that magnolol, applied in vitro to rat neutrophil cells, stimulates an increase in intracellular cytosolic free Ca

in a dose-dependent manner. They observed that this Ca

derives both from intracellular stores and from increased Ca

in¯ux across the cell plasma membrane. Elevated cytosolic free Ca

generally leads to disruption of mitochondrial membrane function, mitochondrial swelling, and the translocation of Cyto c (cytochrome c) out to the cytoplasm [4±6]. Thereafter, the Cyto C evidently

ß 2001 WILEY-LISS, INC.

*Correspondence to: Graduate Institute of Medical Sciences, Taipei Medical University, Taipei 110, Taiwan.

Received 22 February 2001; Revised 9 July 2001; Accepted 18 July 2001

Abbreviations: HUVEC, human umbilical vein endothelial cells;

Cyto c, cytochrome c; MEM, Eagle's minimal essential medium; FCS, fetal calf serum; DMSO, dimethyl sulfoxide; G3PDH, glycerol-3- phosphate dehydrogenase; PARP, poly(ADP-ribose)polymerase;

pNA, p-nitroaniline; U73122, 1-[6-[[(17b)-3-methoxyestra- 1,3,5(10)-trien-17-yl]amino]hexyl]-1 H-pyrrole-2,5-dione; S.E., stan- dard error; BAPTA/AM, 1,2-bis(2-aminophenoxy)ethane-N,N,N

, N

-tetraacetic acid acetoxymethyl ester.

DOI 10.1002/mc.1066

leads to the appearance of caspase-activating pro- teins, the activation of caspase activity, and even- tually the type of cell death known as apoptosis [7,8].

In the present study, we found that 100 mM magnolol, added to cultures of human hepatoma (Hep G2), induced sharp elevations in cytosolic-free Ca

, the activation of Fas-mediated pathways, and the translocation of Cyto c from mitochondria to cytoplasm. Thereafter, widespread apoptosis oc- curred, with marked reduction in cell proliferation.

In Hep G2 cells, we noted that apoptosis was associated with activated caspase activity, while the anti-apoptosis protein bcl-2 was downregulated.

For comparison, human untransformed cells (gingi- val ®broblasts and human umbilical vein endothe- lial cells (HUVEC)) were incubated in culture and treated with 100 mM magnolol; in these noncancer- ous cells, apoptosis and the associated disruption of intracellular organization did not occur. The details of these experiments are described below and shed new light on the molecular mechanisms involved in magnolol-induced apoptosis.

MATERIALS AND METHODS Cell Culture

Three human malignant cell lines (COLO 205, Hep G2, and HT 29) and two types of human primary cells (gingival ®broblasts and HUVEC) were used in this study. COLO 205 (CCL-222; American Type Culture Collection, Rockville, MD) originated from a poorly differentiated human colon adeno- carcinoma. Hep G2 (HB-8065; American Type Culture Collection) was derived from a human hepatoma. HT 29 (HTB-38; American Type Culture Collection) originated from a human colon adeno- carcinoma. Human gingival ®broblasts were har- vested by enzymatic dissociation. HUVEC were harvested from the human umbilical vein by enzymatic dissociation, as previously described [9].

The cells were grown in RPMI 1640 (for COLO 205 and HT 29), Eagle's minimal essential medium (MEM) (for Hep G2 and human ®broblasts), or M199 (for HUVEC), supplemented with 10% fetal calf serum (FCS), penicillin (100 U/mL), streptomy- cin (100 mg/mL), and 0.3 mg/mL glutamine in a humidi®ed incubator (378C, 5% CO

). Magnolol (Pharmaceutical Industry, Technology and Devel- opment Center, Taiwan) was added at the indicated doses in 0.1% dimethyl sulfoxide (DMSO). For control specimens, the same volume of the 0.1%

DMSO without magnolol was added.

Analysis of DNA Fragmentation

As previously described [10], the cells treated with magnolol in 0.1% DMSO or without magnolol (control) cells were seeded onto 100-mm dishes.

The cells were harvested, washed twice with ice-cold

phosphate-buffered saline, resuspended in Tris±

normal saline±EDTA (10 mM Tris-HCl at pH 7.6, 140 mM sodium chloride, and 1 mM EDTA), and lysed in 4 mL of extraction buffer (10 mM Tris-HCl at pH 8.0, 0.1 M EDTA at pH 8.0, 20 mg/mL pancreatic RNase, and 0.5% sodium dodecyl sulfate) at 378C for 2 h. Proteinase K then was added at a

®nal concentration of 100 mg/mL, and the mixture was incubated for another 3 h at 508C. The DNA was extracted twice with equal volumes of phenol and once with chloroform-isoamyl alcohol (24:1 vol/

vol) and then precipitated with 0.1 vol of sodium acetate at pH 4.8 and with 2.5 vol of ethanol at ÿ208C overnight. Finally, it was centrifuged at 13, 000 g for 1 h. Genomic DNA was quantitated, and equal amounts of DNA sample in each lane were electrophoresed in a 2% agarose gel. The DNA was visualized by ethidium bromide staining. Genomic DNA isolated from 10 mM terbina®ne-treated HT 29 cells showing DNA fragmentation was included in this experiment to serve as a positive control.

Flow Cytometry

As previously described [11], the cells were seeded onto 100-mm dishes and grown in MEM supple- mented with 10% FCS. After growing to subcon-

¯uence, the cells were treated with 100 mM magnolol in the presence or absence of various inhibitors in 1% FCS, harvested at 4 h after release with trypsin-EDTA, washed twice with phosphate-buf- fered saline/0.1% dextrose, and ®xed in 70%

ethanol at 48C. The DNA content of the nuclei was determined by staining nuclear DNA with a solution containing propidium iodine (50 mg/mL) and DNase-free RNase (2 U/mL) and measuring the relative DNA content of nuclei using a ¯uores- cence-activated cell sorter. The proportion of nuclei in each phase of the cell cycle was determined using established CellFIT DNA analysis software (Becton Dickinson, San Jose, CA).

Protein Extraction and Western Blot Analysis

Western blot analysis was performed as previously described [12,13]. The protein samples (50 mg per lane) were loaded onto a 10% sodium dodecyl sulfate±polyacrylamide gel, and electrophoresis was performed at a constant 150 V for 3±4 h at 16±188C.

After protein separation, each gel was transferred onto an Immobilon-P membrane. Immunodetec- tion was carried out by probing with appropriate dilutions of speci®c antibodies at room temperature for 2 h. Anti±caspase-3, anti±bcl-2, and anti-bax monoclonal antibodies (Transduction, San Diego, CA); anti±caspase-8 and anti±caspase-9 monoclo- nal antibodies (PharMingen, San Diego, CA); anti±

Cyto c monoclonal antibody (Zymed, San Francisco,

CA); and anti±glycerol-3-phosphate dehydrogenase

(G3PDH) monoclonal antibody (Biogenesis, King-

ston, NH) were used at a dilution concentration of

1:1000. Anti±poly(ADP-ribose)polymerase (PARP) polyclonal antibody (Upstate Biotechnology, Lake Placid, NY) was used at a concentration of 1:250 dilution. The secondary antibodies, alkaline phos- phatase±coupled anti-mouse or anti-rabbit anti- body (Jackson, Westgrove, PA), were incubated at room temperature for 1 h at dilution concentrations of 1:5000 or 1:1000, respectively. The speci®c protein complexes were identi®ed using nitro blue tetrazolium chloride/bromo-chloro-3-indolyl-phos- phate (Kirkegaard Perry Laboratory, Gaithersburg, Maryland). In each experiment, membranes also were probed with anti-G3PDH antibody, to correct for differences in protein loading.

Measurement of Caspase Activity

As previously described [14], caspase activity was measured by using caspase-8 and caspase-9 colori- metric activity assay kits (Chemicon, Temecula, CA). Hep G2 cells were lysed by addition of cell lysis buffer, and the protein concentration was mea- sured. Caspase activity was assayed at 378C in 100 mL of assay buffer containing 25 mg (for caspase-8) or 30 mg (for caspase-9) of the indicated colorimetric peptide. Caspase activity was measured by the release of p-nitroaniline (pNA) from the labeled substrates Ac-IETD-pNA and Ac-LEHD-pNA for cas- pase-8 and caspase-9, respectively, and the free pNA was quanti®ed at 405 nm.

Preparation of Cytosolic Extracts

For study of Cyto c translocation, cytosolic ex- tracts were prepared at various times after the cells were treated with DMSO or magnolol, as previously described [15]. Brie¯y, the Hep G2 cells grown in MEM containing 10% FCS, with or without 100 mM magnolol, were collected by centrifugation at 200 g at 48C for 5 min, washed once with ice-cold phos- phate-buffered saline at pH 7.4, and centrifuged at 200 g for 5 min. The cell pellet was resuspended in extraction buffer (200 mM mannitol, 68 mM sucrose, 50 mM PIPES-KOH at pH 7.4, 50 mM KCl, 5 mM EGTA, 2 mM MgCl

, 1 mM dithiothreitol, and protease inhibitors). After a 30-min incubation period on ice, cells were homogenized with a glass dounce homogenizer (40 strokes) and centrifuged at 14 000 g for 15 min; the supernatants were remo- ved and used for immunoblotting to quantify Cyto c.

Measurement of Cytosolic Free Ca

Concentrations As previously described [16], the Hep G2 cells were seeded onto coverslips and grown in MEM medium supplemented with 10% FCS. After the cells had grown to 60±70% con¯uence, the medium was changed to MEM supplemented with 1% FCS for 24 h and then replaced with fresh MEM supplemented with 1% FCS containing 5 mM fura- 2/AM (Molecular Probes, Eugene, Oregon) and incu-

bated for 45 min at 378C. After washing with Phocal buffer (10 mM HEPES, 125 mM NaCl, 5 mM KCl, 10 mM glucose, 2 mM MgCl

, 0.5 mM NaH

PO

, 5 mM NaHCO

, and 1.8 mM CaCl

) three times, the coverslip was put into a quartz cuvette containing 1.8 mL of Phocal buffer with a magnetic stir bar.

Fluorescence activity was monitored with a ¯uores- cence spectrophotometer (Hitachi F-4500, Tokyo, Japan) at 510 nm, with excitation at 340 and 380 nm.

[Ca

]

was calibrated from the ¯uorescence inten- sity by using the equation [Ca

]

ˆ K

Q[(R ÿ R

)]/

(R

ÿ R)], where R represents the ¯uorescence intensity ratio Fl1/Fl2, in which l1 (340 nm) and l2 (380 nm) are the ¯uorescence detection wavelengths for the ion-bound and ion-free indi- cator, respectively.

The values of F

, F

, R

, and R

were obtained at the end of each experiment by the sequential addition of 10 mM ionomycin (BIOMOL Research Laboratories, Plymouth Meeting, PA) and 50 mM EGTA. Q was the ratio of F

to F

at l2 (380 nm). The dissociation constant (K

) was taken as 240 nM. To study the source of the magnolol-induced elevation of cytosolic free Ca

, the Hep G2 cells were pretreated with 1,2-bis(2- aminophenoxy)ethane-N,N,N

,N

-tetraacetic acid acetoxymethyl ester (BAPTA/AM) or 1-[6-[[(17b)-3- methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1 H-pyrrole-2,5-dione (U73122) for 30 min, followed by treatment with 100 mM magnolol.

Statistics

All data were expressed as the mean values

 standard error (S.E.). Comparisons were subjected to Student's t test or analysis of variance followed by Fisher's least signi®cant difference test. Signi®cance was accepted at P < 0.05.

RESULTS Magnolol Induced Apoptosis in Human Malignant Cell Lines

The occurrence of apoptosis, a natural death

process in cells, most commonly is associated with

arrested mitotic activity, decreased DNA replication,

fragmentation of DNA, and activation of caspase-

type enzymes. Accordingly, we examined COLO

205 and Hep G2 cells treated with magnolol to

determine whether these indications of apoptosis

appeared. At concentrations of 0±50 mM magnolol,

apoptosis was not noted. When the magnolol

concentration was increased to 100 mM, however,

apoptosis was seen in COLO 205 (Figure 1a) and Hep

G2 cells (Figure 1b) but not in cultured human

nontransformed gingival ®broblasts (Figure 1c) and

HUVEC (Figure 1d). Hence, magnolol was used

at 100 mM in all of the remaining experiments in

this study.

Magnolol Activated Caspase-3, Caspase-8, and Caspase-9

Because it has been suggested [17] that apoptosis requires the activation of caspases, we investigated the involvement of caspase activation in magnolol- induced apoptosis in Hep G2 cells by using Western blot analyses. The Hep G2 cells were grown in MEM containing 10% FCS and treated with 100 mM magnolol for various times, as indicated. Figure 2a shows that decreased staining in the procaspase-3 band and degradation of PARP, the substrate for caspase-3, were noted after 36 h of treatment with 100 mM magnolol, indicating that caspase-3 was

activated. To elucidate the apoptotic pathways involved in the activation of caspase-3, we exam- ined the changes of caspase-8 and caspase-9 protein levels in the magnolol-treated Hep G2 cells. After treatment of Hep G2 cells with 100 mM magnolol for 36 h, activation of caspase-8 and caspase-9 was evidenced by degradation of the pro-enzymes of caspase-8 and caspase-9 as well as by the appearance of their cleavage products (Figure 2a).

Magnolol Induced Cyto c Release From Mitochondria It has been shown that activation of caspase- 9 occurs during the release of Cyto c from Figure 1. Electrophoresis of genomic DNA from COLO 205 and

Hep G2 treated with magnolol. Genomic DNA isolated from 10 mM terbina®ne±treated HT 29 cells was included in this experiment to serve as a positive control. A typical DNA ladder pattern associated

with apoptosis was seen in COLO 205 (a) and Hep G2 cells (b) but not

in human gingival ®broblasts (c) or HUVEC (d) treated with 100 mM

magnolol for up to 36 h. P, positive control.

Figure 2. The effect of magnolol on caspase, bcl-2, and bax protein levels and Cyto c translocation from the mitochondria to the cytosol. Whole-cell proteins (a and c) or cytosolic proteins (b) were extracted from cultured Hep G2 cells that had been grown in 10%

FCS and incubated for the indicated times with 0.1% DMSO or 100 mM magnolol in 0.1% DMSO. After electrophoresis, proteins were transferred onto Immobilon-P membranes and then probed with

proper dilutions of speci®c antibodies. Membranes also were probed with anti-G3PDH antibody to correct for any difference in protein loading. Magnolol time dependently induced the activation of caspases (a), the translocation of Cyto c (b), and bcl-2 down- regulation (c). D, 0.1% DMSO; M, 100 mM magnolol in 0.1% DMSO.

The pro-enzymes are indicated by triangles. The cleavage products of

the pro-enzymes are indicated by arrows.

mitochondria [18]. To examine whether this activa- tion occurs in magnolol-induced apoptosis in Hep G2 cells, Cyto c release was monitored at various times after treatment with 100 mM magnolol. Figure 2b shows that magnolol treatment resulted in a signi®cant accumulation of Cyto c in the cytosol fraction of cell extracts. This magnolol-induced elevation of cytosolic Cyto c was noted at 12 h and peaked at 36 h after magnolol treatment. Under the same conditions, caspase activation was not seen until 36 h after magnolol treatment (Figure 2a).

Evidently, translocation of Cyto c occurred in the magnolol-treated Hep G2 cells ®rst, and activation of caspase-8 and caspase-9 and DNA fragmentation followed thereafter.

Magnolol Downregulated Bcl-2 Protein

Proteins of the bcl-2 family also are believed to be involved in the control of apoptosis [19]. Bcl-2 directly or indirectly operates to prevent the release of Cyto c from mitochondria. On the other hand, bax can trigger mitochondria to release Cyto c from mitochondria and thereby initiate apoptosis. Accor- dingly, we examined the changes in bcl-2 protein levels in magnolol-treated Hep G2 cells. Treatment of Hep G2 cells with 100 mM magnolol caused a time-dependent downregulation of bcl-2 protein, as shown by its decreased staining on gels (Figure 2c).

In contrast, bax protein levels were not changed signi®cantly.

Figure 3. Magnolol-induced elevation of cytosolic free Ca

in Hep G2. The cells were grown in MEM supplemented with 1% FCS.

For the measurement of [Ca

]

changes, the fura-2/AM±loaded cells were stimulated with various concentrations (a) or 100 mM (b±

d) of magnolol (arrow). (a) Dose-dependent elevation of [Ca

]

in Hep G2 cells treated with magnolol. (b) Pretreatment with 2 mM EGTA for 100 s did not affect the magnitude of the magnolol-

induced elevation of [Ca

]

in Hep G2 cells. (c) Pretreatment of Hep G2 cells with BAPTA/AM for 30 min reduced the magnitude of the magnolol-induced elevation of [Ca

]

in a dose-dependent manner.

(d) Pretreatment of Hep G2 cells with U73122 for 30 min reduced

the magnitude of the magnolol-induced elevation of [Ca

]

in a

dose-dependent manner. The time points of magnolol and EGTA

administration are indicated by arrows and a triangle, respectively.

Fi gu re 4. Th e ef fe ct of ZB 4 on m ag no lo l-i nd uc ed ap op to sis an d ca sp as e ac tiv ity in H ep G 2 ce lls .T he ce lls w er e gr ow n in M EM su pp le m en te d w ith 1% FC S. Sh ow n ar e re pr es en ta tiv e D N A co nt en tf re qu en cy hi st og ra m s of H ep G 2 ce lls tr ea te d w ith 0. 1% D M SO (a ), 10 0 mM m ag no lo l( b) ,o r 25 0 ng /m L ZB 4 ‡ 10 0 mM m an go lo l( c) .( d) Th e av er ag e G

/G

su bp op ul at io n of 0. 1% D M SO ±t re at ed ve rs us m ag no lo l-t re at ed ve rs us m ag no lo l‡ ZB 4± tr ea te d H ep G 2 ce lls .T re at m en t w ith ZB 4 fo r 4 h su pp re ss ed th e m ag no lo l-i nd uc ed ac tiv at io n of ca sp as e- 8 (e )b ut no t ca sp as e- 9 (f) . Th e ac tiv at io n of ca sp as e- 8 an d ca sp as e- 9 w as ev id en ce d by th e de gr ad at io n of pr o- en zy m es of ca sp as e- 8 an d ca sp as e- 9, as de te ct ed by W es te rn bl ot an al ys is; in cr ea se d ca sp as e ac tiv ity w as ex am in ed by a ca sp as e ac tiv ity as sa y. Va lu es ar e m ea ns  S. E. (n ˆ 3) . C om pa ris on s w er e su bj ec te d to th e St ud en t's t te st . Si gn i® ca nc e w as ac ce pt ed at P < 0. 05 .A st er isk ,m ag no lo l-t re at ed gr ou p di ff er en t fr om ZB 4 ‡ m ag no lo l-t re at ed gr ou p; D ,0 .1 % D M SO ;M ,1 00 mM m ag no lo li n 0. 1% D M SO .

Magnolol Elevated the Cytosolic Free Ca

Concentration

Green and Reed [7] have proposed that the abnormal elevation of cytosolic free Ca

is one of the major occurrences in apoptosis. Previously, Wang and Chen [3] reported that magnolol, applied in vitro to rat neutrophil cells, stimulates an increase in cytosolic free Ca

in a dose-dependent manner. To examine whether magnolol-induced apoptosis is associated with an increase in cytosolic- free Ca

, we compared intracellular fura-2/AM

¯uorescence activity (an indicator of cytosolic free Ca

concentration) in the Hep G2 cells treated with or without 100 mM magnolol. Figure 3a shows an initial rapid spike, with a sustained high [Ca

]

curve in the magnolol-treated Hep G2 cells, which occurred rapidly and in a dose-dependent manner.

Since elevation of [Ca

]

can be caused by either Ca

in¯ux from the external medium or by Ca

released from internal stores, we treated the cells with 2 mM EGTA to chelate the extracellular free Ca

and prevent its entry into cells. Figure 3b shows that pretreatment with 2 mM EGTA for 100 s did not affect the magnitude of magnolol-induced elevation of cytosolic free Ca

; hence, the increase of [Ca

]

is not due to an in¯ux of extracellular Ca

. In contrast, pretreatment of Hep G2 cells with BAPTA/AM, an intracellular Ca

chelator, at 378C for 30 min reduced the magnitude of the magnolol- induced elevation of cytosolic free Ca

(Figure 3c).

Moreover, pretreatment of Hep G2 cells at 378C for 30 min with U73122 (a phospholipase C inhibitor), which inhibits the hydrolysis of phosphatidylino- sitol biphosphate to inositol triphosphate [20], signi®cantly attenuated the magnolol-induced ele- vation of [Ca

]

(Figure 3d).

Fas was Involved in Magnolol-Induced Caspase-8 Activation

It has been suggested that caspase-8 is associated with apoptosis involving a so-called death receptor [21]. Hep G2 cells express both Fas and Fas ligand [22]. Therefore, it is reasonable to speculate that Fas might be involved in caspase-8 activation in magno- lol-treated Hep G2 cells. To address this issue, Hep G2 cells were treated with ZB4, an antibody antagonistic to Fas that binds to the Fas ligand binding site on a Fas receptor and blocks Fas activation. We then examined DNA content fre- quency histograms, caspase levels, and caspase activity. Figure 4a±c shows representative DNA content frequency histograms from DMSO-, mag- nolol-, and ZB4 ‡ magnolol±treated Hep G2 cells. In response to 100 mM magnolol treatment, the G

/G

subpopulation, an indicator of apoptosis, was increased signi®cantly (Figure 4b and d). We also found that pretreatment with ZB4 at a concentra- tion of 250 ng/mL suppressed the magnolol-

induced increase in the G

/G

subpopulation (Fig- ure 4c and d). It is clear that ZB4 partially reversed magnolol-induced apoptosis by 15±20%. ZB4 also inhibited magnolol-mediated caspase-8 activation, as evidenced by Western blot analysis and a caspase activity assay (Figure 4e). In contrast, pretreatment with ZB4 had no effect on magnolol-induced caspase-9 activation (Figure 4f).

Inositol Triphosphate±Mediated Pathway was Involved in Magnolol-Induced Caspase Activation

To study further the effect of elevated [Ca

]

on the development of apoptosis, Hep G2 cells were pretreated with 20 mM U73122 (a phospholipase C inhibitor) for 30 min, followed by 100 mM magnolol treatment. Figure 5a±c shows representative DNA content frequency histograms from DMSO-, mag- nolol-, and U73122 ‡ magnolol±treated Hep G2 cells. Figure 5c and d shows that pretreatment of these cells with U73122 for 30 min resulted in the inhibition of the magnolol-induced increase in the G

/G

subpopulation. The graphs presented as Figure 5e and f show that magnolol-induced activa- tion of caspase-8 and caspase-9 was inhibited substantially by pretreatment of Hep G2 cells with U73122. These ®ndings suggested that the inositol triphosphate±mediated signaling pathway might be involved in the magnolol-induced increase of cytosolic free Ca

, which, in turn, led to apoptosis in Hep G2 cells through activation of caspase-8 and caspase-9.

DISCUSSION

A commonly held view is that uncontrolled cell proliferation in malignant tissues derives from a combination of two circumstances: increased cell multiplication unresponsive to normal control processes and decreased occurrence of the normal process of cell death, that is, apoptosis. Our present state of knowledge, however, offers little insight regarding the intrinsic or extrinsic signals that might govern the opposing processes of cell multi- plication and cell death to maintain normal cell populations. While we were engaged in our inves- tigations of these matters, we discovered that the substance magnolol, derived from a herbal medica- tion, has the potent effect of inducing apoptosis in two human cancer cell types, Hep G2 and COLO 205, but not in human nontransformed cells, such as gingival ®broblasts and HUVEC. In that initial part of our study, we found that even though magnolol, at the concentrations of 0±50 mM, strongly inhib- ited growth and proliferation in these cells, it did not have a perceptible in¯uence on apoptosis.

While continuing this investigation, we found that

magnolol at the higher concentration of 100 mM

used for treatment periods of 36 h or more could

indeed produce the characteristic signs of apoptosis.

Fi gu re 5. Ef fe ct of U 73 12 2 on m ag no lo l-i nd uc ed ap op to sis an d ca sp as e ac tiv ity in H ep G 2 ce lls .T he ce lls w er e gr ow n in M EM su pp le m en te d w ith 1% FC S. Sh ow n ar e re pr es en ta tiv e D N A co nt en tf re qu en cy hi st og ra m s of H ep G 2 ce lls tr ea te d w ith 0. 1% D M SO (a ), 10 0 mM m ag no lo l( b) ,o r 20 mM U 73 12 2 ‡ 10 0 mM m an go lo l( c) .( d) Th e av er ag e G

/G

su bp op ul at io n of 0. 1% D M SO -t re at ed ve rs us m ag no lo l-t re at ed ve rs us m ag no lo l‡ U 73 12 2± tr ea te d H ep G 2 ce lls . Pr et re at m en t of H ep G 2 w ith 20 mM U 73 12 2 pa rt ia lly re ve rs ed m ag no lo l-i nd uc ed ac tiv at io n of ca sp as e- 8 (e )a nd ca sp as e- 9 (f) .T he ac tiv at io n of ca sp as e- 8 an d ca sp as e- 9 w as ev id en ce d by th e de gr ad at io n of pr o- en zy m es of ca sp as e- 8 an d ca sp as e- 9, as de te ct ed by W es te rn bl ot an al ys is; in cr ea se d ca sp as e ac tiv ity w as ex am in ed by a ca sp as e ac tiv ity as sa y. Va lu es ar e m ea ns  S. E. (n ˆ 3) .C om pa ris on s w er e su bj ec te d to St ud en t's t te st .S ig ni ®c an ce w as ac ce pt ed at P < 0. 05 .A st er isk ,m ag no lo l± tr ea te d gr ou p di ff er en tf ro m U 73 12 2 ‡ m ag no lo l± tr ea te d gr ou p. D ,0 .1 % D M SO ;M ,1 00 mM m ag no lo li n 0. 1% D M SO ;U ,U 73 12 2.

These ®ndings show for the ®rst time that magnolol can induce apoptosis in cultured cells.

Apoptosis is a cell-suicide mechanism that requires specialized cellular machinery. A central component of this machinery is a proteolytic system involving caspases, a highly conserved family of cysteine proteinases with speci®c sub- strates [17]. How caspases contribute to this process is not understood fully. It has been suggested that the terminal stages of apoptosis occur through the activation of caspases and that different initiator caspases mediate distinct sets of signals. By Western blot analysis, we showed that 100 mM magnolol induced activation of caspase-8 and caspase-9 in Hep G

cells, as evidenced by decreases in stainable procaspase-8 and procaspase-9, and increases in caspase-8 and caspase-9 enzyme activities (Figure 4e and f). The active forms of caspase-8 and caspase-9 can cleave and activate downstream caspases, such as caspase-3 (Figure 2a), which eventually leads to apoptosis [23].

Caspase-9 is involved in death induced by cytotoxic agents [24,25]. Zou et al. [18] postulated that Cyto c may activate caspases by binding to Apaf-1, which interacts with and activates caspase- 9. Cyto c release sometimes can contribute to Fas- mediated apoptosis by amplifying the effects of caspase-8 to activate downstream caspases [26]. The active caspases, on the other hand, can promote Cyto c release and thereby amplify the signal for apoptosis [27]. In response to 100 mM magnolol, cytosolic Cyto c levels began to increase in about 12 h (Figure 2b), and signi®cant activation of caspase-8 and caspase-9 occurred about 24 h later (Figure 2a). This ®nding suggested that the magno- lol-mediated increase of cytosolic Cyto c could bring about activation of caspase-8 and caspase-9.

Although direct evidence of the involvement of bcl-2 and cytosolic-free Ca

in magnolol-mediated Cyto c release is absent, the data of the present study suggested that downregulation of bcl-2 and elevation of cytosolic free Ca

occurred during

magnolol-mediated release of Cyto c. It has been suggested that bcl-2 family proteins are involved in the regulation of apoptosis through control of the release of Cyto c from mitochondria [19]. Bcl-2 prevents apoptosis by blocking the release of Cyto c from mitochondria [28]. Bax, on the other hand, directly induces Cyto c release from mitochondria and thereby triggers caspase-9 activation [29].

Treatment of isolated rat liver mitochondria with calcium chloride can trigger the release of Cyto c in vitro [30]. The results of the present study showed that elevation of the cytosolic free Ca

concentra- tion and decrease of the bcl-2 protein level induced by 100 mM magnolol treatment in Hep G2 cells seemed to be responsible for stimulating the release of Cyto c. The magnolol-mediated increase in cytosolic free Ca

was released from the intracel- lular sources of calcium through inositol tripho- sphate±mediated pathways but not from Ca

in¯ux across the plasma membrane. This result differs from that for magnolol-treated rat neutro- phils, in which magnolol was found to stimulate Ca

release from internal stores and Ca

in¯ux from extracellular sources across the plasma mem- brane [3]. The discrepancy between these two studies might be due to the differential effects of magnolol on different cell types, such as liver cells versus neutrophils or cancer cells versus nontrans- formed cells.

The presence of the death receptor Fas and its ligand in Hep G2 cells is in accord with the hypothesis that caspase-8 is involved in magnolol- induced apoptosis in Hep G2 cells [22]. Administra- tion of the anti-Fas antibody (ZB4) prevented caspase-8 activation (Figure 4e) and reduced mag- nolol-induced apoptosis by 15±20% (Figure 4d).

Magnolol-mediated caspase-9 activation, however, was not affected by ZB4 treatment (Figure 4f). These results lend support to the idea that a Fas-mediated pathway is involved in magnolol-induced caspase-8 activation and consequent apoptosis in Hep G2 cells. It is noteworthy that surface expression of Fas

Figure 6. Model for magnolol-induced anticancer activity. In response to 100 mM magnolol administration, Fas was activated and Cyto c (solid circles) was translocated from the mitochondria to the cytoplasm through elevation of the cytosolic free Ca

con- centration and bcl-2 downregulation. Caspase-8 was activated by

both Fas activation and Cyto c release from mitochondria, whereas

caspase-9 was activated by Cyto c release. The active forms of

caspase-8 and caspase-9 can cleave and activate downstream

caspases, such as caspase-3, which, in turn, induce apoptosis.

was heterogeneous in malignant cell lines [31].

Therefore, it seems that activation of the Fas- mediated pathway does not always participate in magnolol-induced apoptosis. Whether magnolol activated Fas directly or promoted the action of a Fas ligand, which, in turn, activated Fas, remains to be determined.

Based on the results of the present study, we propose a model of the molecular mechanisms of magnolol-induced apoptosis in malignant cell lines, as shown in Figure 6. Although animal studies of magnolol-mediated anti-tumor action are ongoing, the ®ndings from our previous study of magnolol's anti-cancer effect and the present in vitro studies strongly support the potential applications of magnolol in the treatment of human cancer.

ACKNOWLEDGMENTS

We thank Professor Winton Tong (University of Pittsburgh, PA) and Professor Ling-Ru Lee (University of California at Berkeley, Berkeley, CA) for critical review of the manuscript, Dr. Chien-Huang Lin (Taipei Medical University, Taipei, Taiwan) for valu- able discussion, and Dr. How Tseng (Taipei Medical University, Taipei, Taiwan) for editorial assistance.

This work was supported by the foundation of Jin Lung Yen and Sagittarius Life Science Corp.

REFERENCES

1. Akriviadis EA, Llovet JM, Efremidis SC, et al. Hepatocellular carcinoma. Br J Surg 1998;85:1319±1331.

2. Konoshima T, Kozuka M, Tokuda H, et al. Studies on inhibitors of skin tumor promotion. IX. Neolignans from Magnolia of®cinalis. J Nat Prod 1991;54:816±822.

3. Wang J-P, Chen C-C. Magnolol induces cytosolic-free Ca

elevation in rat neutrophils primarily via inositol trisphosphate signalling pathway. Eur J Pharmacol 1998;

352:329±334.

4. Bernardi P, Broekemeier KM, Pfeiffer DR. Recent progress on regulation of the mitochondrial permeability transition pore: A cyclosporin-sensitive pore in the inner mitochondrial membrane. J Bioenerg Biomembr 1994;26:509±517.

5. Zoratti M, Szabo I. Electrophysiology of the inner mitochon- drial membrane. J Bioenerg Biomembr 1994;26:543±553.

6. Zoratti M, Szabo I. The mitochondrial permeability transi- tion. Biochim Biophys Acta±Rev Biomembr 1995;1214:

139±176.

7. Green DR, Reed JC. Mitochondria and apoptosis. Science 1998;281:1309±1312.

8. Hengartner MO. Apoptosis: Death cycle and Swiss army knives. Nature 1998;391:441±442.

9. Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived form umbilical veins:

Identi®cation by morphologic and immunologic criteria. J Clin Invest 1973;52:2745±2756.

10. Ho Y-S, Lee H-M, Mou T-C, Wang Y-J, Lin J-K. Suppression of nitric oxide±induced apoptosis by N-acetyl-

-cysteine through modulation of glutathione, bcl-2, and bax protein levels. Mol Carcinog 1997;19:101±113.

11. Lai K, Wang H, Lee W-S, Jain MK, Lee M-E, Haber E.

Mitogen-activated protein kinase phosphatase-1 in rat arterial smooth muscle cell proliferation. J Clin Invest 1996;98:1560±1567.

12. Lee W-S, Jain MK, Arkonac BM, et al. Thy-1, a novel marker for angiogenesis upregulated by in¯ammatory cytokines.

Circ Res 1998;82:845±851.

13. Lin C-H, Sheu S-Y, Lee H-M, et al. Involvement of protein kinase c-gamma in IL-1 b, b-induced cyclooxygenase-2 expression in human pulmonary epithelial cells. Mol Pharmacol 2000;57:36±43.

14. Deveraux QL, Leo E, Stennicke HR, Welsh K, Salvesen GS, Reed JC. Cleavage of human inhibitor of apoptosis protein XIAP results in fragments with distinct speci®cities for caspases. EMBO J 1999;18:5242±5851.

15. Bossy-Wetzel E, Newmeyer DD, Green DR. Mitochondrial cytochrome c release in apoptosis occurs upstream of DEVD-speci®c caspase activation and independently of mitochondrial transmembrane depolarization. EMBO J 1998;17:37±49.

16. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca

indicators with greatly improved ¯uorescence proper- ties. J Biol Chem 1985;260:3440±3450.

17. Thornberry NA, Lazebnik Y. Caspases: Enemies within.

Science 1998;281:1312±1316.

18. Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein homologous to C. elegans CED-4, partici- pates in cytochrome c±dependent activation of caspase-3.

Cell 1997;90:405±413.

19. Reed JC. Bcl-2 family proteins. Oncogene 1998;17:3225±

3236.

20. Yule DI, Williams JA. U73122 inhibits Ca

oscillations in response to cholecystokinin and carbachol but not to JMV- 180 in rat pancreatic acinar cells. J Biol Chem 1992;267:

13830±13835.

21. Ashkenazi A, Dixit VM. Death receptors: signaling and modulation. Science 1998;281:1305±1308.

22. Jiang S, Song M-J, Shin E-C, Lee M-O, Kim S-J, Park JH.

Apoptosis in human hepatoma cell lines by chemother- apeutic drugs via Fas-dependent and Fas-independent pathways. Hepatology 1999;29:101±110.

23. Li P, Nijhawan D, Budihardjo I, et al. Cytochrome c± and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade. Cell 1997;91:479±

24. Kuida K, Haydar TF, Kuan CY, et al. Reduced apoptosis and 489.

cytochrome c±mediated caspase activation in mice lacking caspase 9. Cell 1998;94:325±337.

25. Hakem R, Hakem A, Duncan GS, et al. Differential requirement for caspase 9 in apoptotic pathways in vivo.

Cell 1998;94:339±352.

26. Kuwana T, Smith JJ, Muzio M, Dixit V, Newmeyer DD, Kornbluth S. Apoptosis induction by caspase-8 is ampli®ed through the mitochondrial release of cytochrome c. J Biol Chem 1998;273:16589±16594.

27. Chen Q, Gong B, Almasan A. Distinct stages of cytochrome c release from mitochondria: Evidence for a feedback ampli®cation loop linking caspase activation to mitochon- drial dysfunction in genotoxic stress induced apoptosis. Cell Death Differ 2000;7:227±233.

28. Yang J, Liu X, Bhalla K, et al. Prevention of apoptosis by bcl- 2: release of cytochrome c from mitochondria blocked.

Science 1997;275:1129±1132.

29. Jurgensmeier JM, Xie Z, Deveraux Q, Ellerby L, Bredesen D, Reed JC. Bax directly induces release of cytochrome c from isolated mitochondria. Proc Natl Acad Sci USA 1998;95:

4997±5002.

30. Kantrow SP, Piantadosi CA. Release of cytochrome c from liver mitochondria during permeability transition. Biochem Biophys Res Commun 1997;232:669±671.

31. Moller P, Koretz K, et al. Leithauser F. Expression of APO-1 (CD 95), a member of the NGF/TNF receptor superfamily, in normal and neoplastic colon epithelium. Int J Cancer 1994;

57:371±377.