Honokiol Induces Calpain-Mediated Glucose-Regulated Protein-94 Cleavage and Apoptosis in Human Gastric Cancer Cells and Reduces Tumor Growth

Background Honokiol, a small molecular weight natural product, has been shown to possess potent anti-neoplastic and anti-angiogenic properties. Its molecular mechanisms and the ability of anti-gastric cancer remain unknown. It has been shown that the anti-apoptotic function of the glucose-regulated proteins (GRPs) predicts that their induction in neoplastic cells can lead to cancer progression and drug resistance. We explored the effects of honokiol on the regulation of GRPs and apoptosis in human gastric cancer cells and tumor growth. Methodology and Principal Findings Treatment of various human gastric cancer cells with honokiol led to the induction of GRP94 cleavage, but did not affect GRP78. Silencing of GRP94 by small interfering RNA (siRNA) could induce cell apoptosis. Treatment of cells with honokiol or chemotherapeutics agent etoposide enhanced the increase in apoptosis and GRP94 degradation. The calpain activity and calpain-II (m-calpain) protein (but not calpain-I (µ-calpain)) level could also be increased by honokiol. Honokiol-induced GRP94 down-regulation and apoptosis in gastric cancer cells could be reversed by siRNA targeting calpain-II and calpain inhibitors. Furthermore, the results of immunofluorescence staining and immunoprecipitation revealed a specific interaction of GRP94 with calpain-II in cells following honokiol treatment. We next observed that tumor GRP94 over-expression and tumor growth in BALB/c nude mice, which were inoculated with human gastric cancer cells MKN45, are markedly decreased by honokiol treatment. Conclusions and Significance These results provide the first evidence that honokiol-induced calpain-II-mediated GRP94 cleavage causes human gastric cancer cell apoptosis. We further suggest that honokiol may be a possible therapeutic agent to improve clinical outcome of gastric cancer.


INTRODUCTION
Gastric cancer is the second most common cause of cancer death in the world [1]. Almost two-thirds of the cases occur in developing countries and 42% in China alone [1,2]. The molecular targets and mechanisms underlying poor prognosis are not well understood. The treatment of locally advanced gastric cancer remains a major challenge. Despite recent advances in treatment, the clinical outcome for gastric cancer patients remains poor. Apart from surgery, the role of adjuvant therapy remains unproven [2,3]. Thus the need to identify potential novel therapeutic and chemopreventive agents is obvious.
Honokiol, an active component isolated and purified from the Magnolia officinalis, has been demonstrated to possess the effects of anti-oxidation [4], against lipid peroxidation [5] and antiinflammatory in vitro and in vivo [6,7]. Honokiol has also been shown to be a systemically available and non-toxic inhibitor of angiogenesis [8]. The recent studies reported that honokiol induces caspase-dependent apoptosis in B-cell chronic lymphocytic leukemia cells and overcomes conventional drug resistance in human multiple myeloma by induction of caspase-dependent andindependent apoptosis [9,10]. Although a full assessment of the clinical potential of the compounds is not yet possible, the pharmacokinetics of honokiol has been thoroughly investigated, revealing that honokiol is available upon clinical cancer therapy. More natural products containing a variety of therapeutic compounds useful in preventing the development of malignancies continue to be discovered; however, very little is known about their underlying mechanisms of action or their molecular target.
Glucose-regulated protein (GRP)94 is a most abundant glycoprotein in endoplasmic reticulum (ER) and only be identified in vertebrate. Over-expression antisense and ribozyme approaches in tissues culture systems directly demonstrated that GRP78 and GRP94 could protect cells against death [11][12][13]. The protective function of the GRPs has also been observed in resistance to radiation in cervical cancer [14]. The anti-apoptotic function of the GRPs also predicts that their induction in neoplastic cells can lead to cancer progression and drug resistance [15][16][17][18]. Pathological conditions such as tumor growth and malignancy have been suggested to correlate with cytoprotective protein GRP94 over-expression [19]. In metastatic malignancies model, a significant efficacy of the GRP94-based gene/immunotherapy strategy was shown when it was combined with radiation therapy [20]. The ER plays a direct role in activating a subset of caspase during activation of apoptosis that occurs during ER stress [21]. On the other hand, calpains are a family of Ca 2+ -dependent intracellular cysteine proteases. Ubiquitously expressed calpain-I (m-calpain) and calpain-II (m-calpain) proteases are implicated in development of apoptosis. A recent study has shown that ubiquitous calpains promote caspase-12 and JNK activation during ER stressinduced apoptosis [22]. It has also been indicated that GRP94 with Ca 2+ -binding and anti-apoptotic properties is a proteolytic target of calpain during etoposide-induced apoptosis [23]. Moreover, in several experimental models of apoptosis, it has been shown that the amino-terminal calpain inhibitory unit of calpastatin can be cleaved by caspases, suggesting the cleavage is essential for the regulation of calpain activity during cell death [24][25][26][27][28]. Caspase-7, which is recruited to the ER in stressed cells, may likewise cleave and activate caspase-12 [29][30][31]. The effects of honokiol on the GRPs-related signaling and apoptosis still remain unknown. In the present study, we explored the molecular mechanisms of honokiol on human gastric cancer cell apoptosis and tumor growth in vitro and in vivo. Unexpectedly, we found that GRP94 undergoes specific proteolytic cleavage by calpain during honokiol-induced apoptosis. These observations may give evidences that honokiol is a possible therapeutic agent to improve clinical outcome of gastric cancer.

Kinetics of GRP94 proteolytic cleavage in human gastric cancer cell lines treated with honokiol
We firstly examined the effects of honokiol on the expressions of GRP94 and GRP78 in various human gastric cancer cell lines. Honokiol markedly decreases the levels of GRP94 in a dose-and time-dependent manner, but GRP78 levels are not affected (Figs. 1 and 2). Kinetics of honokiol-induced GRP94 cleavage in various human gastric cancer cells were shown in Fig. 2. MKN45 or SCM-1 cells were more resistant to honokiol-induced responses than AGS or N87 cells; the levels of GRP94 were reduced to about 50% for twice as much time. The levels of GRP78 proteins remain unchanged in all four gastric cancer cell lines. Moreover, there are little GRP94 protein expressions in normal mouse gastric epithelium tissue and human endothelial cells as compared with gastric cancer cells (Fig. 1C).
Honokiol induces GRP94 cleavage-associated apoptotic response As shown in Fig. 3, honokiol increased the poly(ADP-ribose) polymerase (PARP) cleavage and DNA damage inducible gene CHOP/GADD153 (a protein has been identified to mediate ER stress-induced apoptosis) levels in MKN45 cells in a dose-and time-dependent manner. Honokiol 20 and 40 M also triggered the expressions of cleaved caspase-12 and caspase-7 (p20), which the magnitude of increase was proportional to the timing (Fig. 3B). Moreover, to understand whether GRP94 cleavage was involved in the human gastric cancer cell apoptosis, apoptosis was detected in GRP94-siRNA-transfected cells. Silencing of GRP94 by siRNA induced cell apoptosis (Fig. 4A) and decreased GRP94 protein expression ( Fig. 4B-a). We also compared the effects of honokiol on apoptosis and GRP94 degradation in human gastric cancer cells with chemotherapeutics agent etoposide. The results showed that treatment of cells with honokiol or etoposide enhanced the increase in apoptosis (Fig. 4A) and GRP94 degradation ( Fig. 4B-b). When cells were simultaneously treated with 40 M etoposide and 20 M honokiol, a larger increase in GRP94 degradation than they did was shown ( Fig. 4B-b); these results imply that the combining of honokiol with other anticancer drugs may be a potential therapeutic strategy.

Calpain is required for honokiol-mediated cell apoptosis
We next determined whether the activity of calpain (calciumdependent thiol proteases) would be induced by honokiol in four gastric cancer cell lines. As shown in Fig. 5A, honokiol 20 M increased calpain activity in a time-dependent manner, which started to increase at 15 min, and peaked at 60 min, and then dropped down to a lower level at 4 h. Cells remained adherent over the time course, with no loss of viability (data not shown). In Fig. 5B, the results showed that the increase of calpain activity in response to honokiol (20 and 40 M) for 60 min in four gastric cancer cell lines. Furthermore, calpain inhibitors, Nacetyl-leu-leu-norleucinal (ALLN), N-acetyl-leu-leu-methioninal We further tested whether decreased calpain activity could compromise the ability of honokiol on apoptosis induction. As shown in Fig. 6A, honokiol (40 M) induced cell apoptosis in SCM-1 cell, which could be reversed by calpain inhibitors (ALLN or ALLM). Moreover, honokiol enhanced calpain-II protein but not calpain-I protein expressions in SCM-1 cells (Fig. 7A). Using a siRNA approach to knockdown calpain-II activity led to a significant abatement of honokiol (20 M)-induced apoptosis in human gastric cancer cells after 4 h treatment (Fig. 6B). SiRNA-calpain-I did not affect honokiol-induced apoptosis (Fig. 6B).

Localization and interaction of calpain-II and GRP94 following honokiol treatment
The next step was to elucidate the role of calpain-II in honokiolinduced GRP94 degradation. In laser confocal microscopic study, the localization of calpain-II protein was determined by green fluorescence, whereas the localization of GRP94 was determined by red fluorescence. As shown in Fig. 7B, both calpain-II and GRP94 were detected in cytoplasm with co-localization in human gastric cancer cells. The fluorescence of calpain-II was gradually increased, but fluorescence of GRP94 was gradually decreased in human gastric cancer cells under honokiol treatment. To further confirm the results from immunocytochemical staining that    8B and 8C). Furthermore, the specific cleavage of GRP94 by calpain could be observed using cell lysates prepared from human gastric cancer cells (data not shown).
We next investigated whether the calpain activation is required for honokiol-induced cell apoptosis. In Fig. 9, the increases of calpain-II protein expression and GRP94 degradation and caspase-7 and caspase-12 activation induced by honokiol (40 M) were prevented by pharmacological calpain inhibitors and transient transfection of siRNA-calpain-II in SCM-1 cells.

Honokiol attenuates tumor GRP94 over-expression and tumor growth in nude mice
Nude mice were inoculated with 4610 6 undifferentiated adenocarcinoma cells MKN45 and treated with honokiol 0.5 and 1.5 mg/kg or vehicle when tumor became evident. Immunohistological and Western blotting analysis showed that GRP94 over-expression and accumulated in tumor region, but not in normal gastric tissue (Figs. 10A-a and 10A-b). Honokiol markedly decreased the accumulation of GRP94 in tumors as compared with vehicle control (Figs. 10A-a and 10A-b). The expression of GRP78 was not affected (data not shown). To determine whether honokiol could suppress tumor growth in vivo, solid tumors were established in mice and the anti-tumor effect of repeatedly injected honokiol was studied. As shown in Fig. 10B, the administration of honokiol 0.5 and 1.5 mg/kg showed a significant anti-tumor activity.

DISCUSSION
We reported here for the first time that calpain, a nonlysosomal Ca 2+ -activated cysteine protease, especially calpain-II, plays a key role in the cleavage of GRP94, but GRP78 is not affected, and in the regulation of apoptosis induced by honokiol in human gastric cancer cells. In particular, we provided functional evidences that the cleavage of GRP94 upon honokiol treatment acts as a therapeutic benefit, inhibiting tumor growth in mouse model of human gastric carcinoma. To our knowledge, this is the first report to show that honokiol can manipulate calpain-inducible chaperone protein GRP94 degradation being the target during apoptosis.
Studies from multiple laboratories pointed to the ER as a target compartment by therapeutic agents implicated in apoptosis execution. Cytosolic Ca 2+ has been implicated as a second messenger of proapoptosis involved in both triggering apoptosis and regulating caspases or calpains [32][33][34]. A recent study has shown that calpain is an important mediator of resveratrol (a natural plant polyphenol)-induced apoptosis in breast cancer [35]. They found that resveratrol can cause an increase in intracellular Ca 2+ , and activates calpain-mediated apoptosis, leading to the degradation of plasma membrane Ca 2+ -ATPase isoform 1 and fodrin in caspase-3 deficient MCF-7 cells. Honokiol has also been reported to induce Ca 2+ mobilization in rat cortical neurons and human neuroblastoma SH-SY5Y cells, presumably through the activation of phospholipase C and IP 3 receptors [36]. In the present study, we found that honokiol is capable of increasing calpain activity and calpain-II protein expression but not calpain-I during honokiol-induced gastric cancer cell apoptosis. Moreover, calpain inhibitors and silencing of calpain-II by siRNA significantly reversed the honokiol-induced apoptosis. These findings imply that Ca 2+ -triggered calpain-II activation may play a potential role in honokiol-induced apoptosis.
GRP94/gp96 is the ER resident member of the HSP90 family. GRP94 plays an important role in maintaining cellular homeostasis. This conventional concept of GRP94 as protein folding chaperone is updated by the discoveries that GRPs promote tumor proliferation, metastasis, drug resistance, and immunotherapy, which have major clinical implications in the prognosis and treatment of cancer [3]. GRP94 has been demonstrated to be associated with tumorigenicity in cancer cell lines, rodent tumor models and human cancer biopsies [37][38][39]. This is consistent with the findings that the induction of GRPs causes a protective function as the survival responses to nutrient starvation, acidosis, and hypoxia conditions, which are common in poorly vascularized solid tumors [11,40]. It has also been appeared that most of the GRPs in tumor cells is engaged in multi-chaperone complexes, while it is not in normal cells [41]. Vaccination of mice with tumor-derived stress proteins, GRP94, can elicit anti-tumor immune responses, yielding a marked suppression of tumor growth and metastasis. The activity of Hsp90 inhibitors has been well validated in preclinical breast cancer models, both in singleagent studies and in combination with conventional chemotherapy [41]. Moreover, it has been suggested that GRP94 is a physiological substrate for calpain [23]. Calpain has been demonstrated to be activated at the ER membrane, where it interacts with GRP94, resulting in its specific proteolytic cleavage as the cells undergo apoptosis triggered by etoposide [23]. In current study, we found that honokiol-induced GRP94 cleavage and apoptosis in human gastric cancer cells can be completely reversed by calpain inhibitors and silencing of calpain-II by siRNA. Caspase inhibitors at high dose partially reversed the honokiol-induced GRP94 cleavage. Pretreatment of cells with the cathepsin B and L inhibitors and proteasome inhibitor was unable to block honokiol-induced GRP94 cleavage. We also demonstrated that silencing of GRP94 by siRNA can induce gastric cancer cell apoptosis. Furthermore, the results of immunocytochemical staining and immunoprecipitation revealed a specific interaction of GRP94 with calpain-II in gastric cancer cells following honokiol treatment. The specific cleavage of GRP94 by calpain could also be observed using cell lysates prepared from human gastric cancer cells. These findings, therefore, suggest that calpain-II-mediated GRP94 cleavage plays an important role in honokiol-triggered gastric cancer cell apoptosis.
Calpains and caspases are two families of cysteine proteases that they are involved in regulating pathological cell death [22,31]. These proteases share several death-related substrates including caspases themselves, cytoskeletal proteins, Bax, and Bid [42]. Calpain-mediated proteolysis proceeds in a limited manner but does not require a specific amino acid residue like that of caspases. Although both calpain and caspase have been proposed to play important roles in regulating pathological cell death, the interac- tions of these two families of proteases under pathological conditions are not clear. In the present study, knockdown and pharmacological inhibitors approaches have contributed significantly to our knowledge of calpain biology, particularly with respect to its specific function on cell apoptosis, which is possible that caspases 7 and 12 are downstream from calpain in mediating honokiol-induced gastric cancer cell apoptosis.
Natural product drugs have been suggested to play a dominant role in pharmaceutical care [43]. Natural products are one of the important sources of potential cancer chemotherapeutic and chemopreventive agents [43]. Honokiol has been widely used in the traditional Chinese and Japanese medicine for several thousand years, mainly, for the treatment of anti-thrombocytic, anti-bacterial, anti-inflammatory, and anxiolytic effects. Previous reports have demonstrated that honokiol is also possessing potent anti-neoplastic and anti-angiogenic properties [6,19,38]. However, the precise molecular mechanism of exhibited anti-tumor activity by honokiol is not well understood. Thus, the results of this study provide evidences for the anti-tumor activity of honokiol in gastric cancer, and more importantly, the molecular basis for its effect. We found that honokiol induces the activation of calpain, GRP94 cleavage, and apoptosis in human gastric cancer cells. Moreover, after administration of honokiol in nude nice implanted with MKN45 gastric tumor cells showed significant anti-tumor activity.
This preclinical study can serve as a framework and represents a promising novel targeted approach. In addition, the natural compounds have been shown to combine with conventional cytotoxic agents may be clinical beneficial with anticancer drugs. This advantage plus the emerging evidence of its anti-tumor mechanisms make honokiol ongoing clinical application for being an effective and safe anti-tumor agent.

Cells, Culture Conditions and Reagents
Human cell lines including Caucasian gastric cancer cell lines (AGS, a moderately-poorly differentiated adenocarcinoma cell line, and N87, a well differentiated carcinoma cell line) and Asian gastric cancer cell lines (MKN45 and SCM-1, the undifferentiated adenocarcinoma cells) were obtained from cell bank in cancer center of Taipei veterans general hospital (Taiwan). Cells were maintained in RPMI 1640 medium containing 10% heatinactivated FCS (Life Technologies) and streptomycin/penicillin (Life Technologies) in a humidified 5% CO 2 atmosphere. Honokiol was obtained from Wako Chemical Company (Japan), and its purity was determined to be a minimum of 99% by highperformance liquid chromatography.

Western Blot Analysis and Immunoprecipitations
Whole cell lysates were prepared as described previously [44]. Proteins were separated by pre-cast 8-20% SDS-polyacrylamide gel electrophoresis, and then electrophoretically transferred from the gel onto polyvinylidene difluoride membranes. After blocking, blots were incubated with, anti-GRP94, anti-GRP78, anti-caspase 12, anti-caspase 7, anti-PARP, anti-GADD153 (Santa Cruz Biotechnology), anti-calpain-I (m-calpain), anti-calpain-II (m-calpain) (Santa Cruz Biotechnology), and anti-b actin (Sigma) antibodies in PBS within 0.1% Tween 20 for 1 h followed by three 10 min washes in PBS within 0.1% Tween 20. The membranes were then incubated with horseradish peroxidaseconjugated secondary antibodies for 60 min. In immunoprecipitation, proteins (500 g) were incubated with specific antibodies and immobilized onto protein A-Sepharose beads. Beads were washed extensively with Bacco9s immunoprecipitative buffer, boiled, and microcentrifuged. Input 10% of cell lysate for IP and 50 g proteins were analyzed by Western blotting. Detection was performed with Western blotting reagent ECL (Amersham), and chemiluminescence was exposed by the Kodak X-Omat films.

Calpain Activity Assays
Suc-Leu-Leu-Val-Tyr-AMC is a calpain protease substrate. Quantitation of 7-amino-4-methylcoumarin (AMC) fluorescence permits the monitoring of enzyme hydrolysis of the peptide-AMC conjugate and can be used to measure enzyme activity. Cells were prepared and treated on 24-well Corning/Costar plates. Prior to addition of inhibitors cells were loaded with 40 M Suc-Leu-Leu-Val-Tyr-AMC (Biomol) and treated with honokiol for indicated timing at 37uC in a humidified 5% CO 2 incubator. Proteolysis of the fluorescent probe was monitored using a fluorescent plate reading system (HTS-7000 Plus Series BioAssay, Perkin Elmer) with filter settings of 360620 nm for excitation and 460620 nm for emission.

SiRNA and Transfection Assays
The siRNA duplexes specific for the inhibition of calpain-I (mcalpain) and calpain-II (m-calpain) expressions in human cells were obtained from Santa Cruz Biotechnology: calpain-I, catalog no. sc-29885, a pool of 3 target-specific 20-25 nt siRNAs; calpain-

Immunofluorescence and Laser Scanning Confocal Microscopy
Cells were washed twice with PBS, fixed in 4% paraformaldehyde for 30 min and then blocked by incubation in 1% bovine serum albumin in PBS. Primary antibodies as indicated were applied to the slides at a dilution of 1:500 and incubated at 4uC overnight. The samples were treated with FITC-conjugated or TRITCconjugated secondary antibodies (Sigma). The FITC-labeled or TRITC-labeled cells were then analyzed by fluorescence microsco-py. For laser scan microscopy, immunofluorescence-labeled cells were analyzed using an inverted laser scanning microscope (Zeiss LSM 410 invert, Germany), equipped with both argon ion (488 nm) and HeNe (543 nm) lasers. Co-localization of two labeled antigens was detected as a single image when the images from both channels were overlaid. Confocal images were background-subtracted and merged using the Confocal Assistant software program, and processed with Adobe Photoshop software.

Annexin-V FITC and PI Double Staining
The annexin V/propidium iodide (BD Clontech) was used to quantify numbers of apoptotic cells as described previously [44]. Cells were washed twice with PBS and stained with annexin V and PI for 20 min at room temperature. The level of apoptosis was determined by measuring the fluorescence of the cells by flow cytometer (Becton Dickinson). Data acquisition and analysis were performed by the CellQuest program (Becton Dickinson).

Animals
Animal experiments were carried out according to the guidelines issued by the Committee of Animal Experiments, National Chung Hsing University, Taichung, Taiwan. Male BALB/c nude mice (nu/nu) were purchased from NLAC Taiwan, Inc., Taipei, Taiwan. The mice were bred and maintained under specific pathogen-free conditions, provided with sterilized food and water ad libitum, and housed in a barrier facility with 12 h light/dark cycles. Experimental procedures were done on 6-week-old mice. The animals were euthanized when they appeared moribund.

Tumor Growth Assay
Cell tumorigenicity was examined in 6-week-old male BALB/c nude mice. Briefly, 4610 6 of cells (MKN45) in an exponential growth phase were suspended in 1 ml culture medium with 10% fetal bovine serum. A single-cell suspension of MKN45 was subcutaneously inoculated into the dorsal side of the mice. Tumors were established for 14 days after MKN45 cells inoculation and followed by treatment with vehicle, 0.5 and 1.5 mg/kg honokiol every day for 10 days (7 mice/group). The volume of the implanted tumor in dorsal side of mice was measured twice a week with a caliper, using the formula V = (LW 2 ) p/6: where V, volume (mm 3 ); L, biggest diameter (mm); W, smallest diameter (mm).

Immunohistochemistry
The expression of GRP94 in mouse solid tumor was examined by immunohistochemistry. Thin tumor sections (5 mm) were prepared, fixed with cold acetone, blocked with 3% hydrogen peroxidase, and stained with for primary GRP94 antibody.

Statistical Analyses
The values given in this study are presented as mean6SEM. All analyses were performed by analysis of variance followed by a Fisher's least significant difference test. P value of less than 0.05 was viewed as statistical significance.