Am J Cancer Res 2016;6(6):1331-1344www.ajcr.us /ISSN:2156-6976/ajcr0028451

Original ArticleEMMPRIN, SP1 and microRNA-27a mediate physcion 8-O-β-glucopyranoside-induced apoptosis in osteosarcoma cells

Zhaohong Wang, Huilin Yang

Department of Orthopaedic Surgery, The First Affiliated Hospital of Soochow University, No 188, Shizi Street, Suzhou 215006, China

Received March 17, 2016; Accepted April 1, 2016; Epub June 1, 2016; Published June 15, 2016

Abstract: Physcion 8-O-β-glucopyranoside (PG), the main active ingredient of Rumex japonicus, induces apoptosis and causes cell cycle arrest in human lung cancer cells. However, its anti-tumor effects are not fully understood. In this study, we explored the mechanisms underlying PG induced apoptosis in the osteosarcoma cell line MG-63. Our results showed that PG exerted anti-proliferative effects and induced apoptosis in MG-63 cells via the intrinsic mi-tochondrial pathway, accompanied by loss of mitochondrial membrane potential (MMP) and cytochrome C release from the mitochondria. In addition, physcion treatment significantly inhibited extracellular matrix metalloproteinase inducer (EMMPRIN) expression in MG-63 cells, in a dose-dependent manner; meanwhile, EMMPRIN protein over-expression markedly reduced PG-induced apoptosis. Moreover, our findings suggested that the modulatory effects of PG on EMMPRIN were due, at least in part, to regulation of an ROS-miR-27a/ZBTB10-Sp1 transcription factor pathway.

Keywords: Physcion 8-O-β-glucopyranoside, osteosarcoma, EMMPRIN, Sp1, miR-27a, ROS

Introduction

Osteosarcoma (osteogenic sarcoma, OS), the most common malignant bone tumor in chil-dren and young adults, accounts for 6% of all childhood cancers [1]. The etiology of this neo-plasm is complex, and multiple risk factors for OS occurrence have been identified, including genetic and environmental factors [2]. However, early diagnosis is rare due to lack of specific early symptoms and signs. Average diagnostic delay for OS is 9 weeks, and tendonitis has been found as the most common misdiagnosis [3]. Modern treatments, such as surgery and chemotherapy used as mono- or combination therapies, have been improved, but prognosis remains poor for patients with osteosarcoma [4, 5]. Furthermore, surgery cannot stop tumor metastasis, and chemotherapy is limited by cancer cell resistance and various side effects. Currently, traditional Chinese medicine (TCM) is commonly administered to cancer patients as an adjunct to conventional therapies, to improve the quality of life by alleviating symp-toms and side effects [6]. Indeed, multiple active gradients of TCM inhibit tumor cell prolif-

eration and induce tumor cell apoptosis [7, 8]. Therefore, developing novel agents for OS treat-ment can help improve the clinical outcome.

Extracellular matrix metalloproteinase inducer (EMMPRIN), a transmembrane glycosylated member of the immunoglobulin superfamily molecules expressed on the cell surface of most tumor cells, was firstly found to increase tumor invasion by inducing matrix metallopro-teinase (MMP) synthesis by the surrounding stromal cells [9]. Other studies showed that EMMPRIN promotes survival, invasion and metastasis in tumor cells through multiple pathways and mechanisms, including loss of function of the tumor suppressor p53 [10], upregulated vascular endothelial growth factor (VEGF) [11, 12], disruption of the growth-modu-lating factor transforming growth factor-β1 (TGF-β1) [13], and regulation of the urokinase-type plasminogen activation (uPA) system of serine proteases [14]. Interestingly, multiple studies suggested EMMPRIN mediates cancer cell apoptosis in a hyaluronan-dependent man-ner via the PI3K and Erk1/2 pathways [15, 16]. Therefore, targeting EMMPRIN may provide a

PG induce apoptosis in osteosarcoma cell

1332 Am J Cancer Res 2016;6(6):1331-1344

mechanistic basis for apoptosis-inducing thera-pies. In the case of osteosarcoma, high levels of EMMPRIN are positively correlated with the clinicopathological degree, and negatively cor-related with patient survival [17, 18]. However, whether EMMPRIN plays a role in OS cell apop-tosis remains to be elucidated.

Rumex japonicus Houtt, a perennial herbal plant belonging to the family Polygonaceae widely distributed in China (known as Yang-Ti, in Chinese), has been used as antimicrobial, pur-gative, anti-inflammatory and anti-tumor agent in the folk medicine for many years [19-21]. Interestingly, a recent research showed that one of its main active ingredients, physcion 8-O-β-glucopyranoside (PG), induces apoptosis and causes cell cycle arrest in the human lung cancer cell line A549 [22]. However, the under-lying mechanisms of PG induced apoptosis remain ununderstood. In this study, the anti-tumor properties of PG were assessed, examin-ing its effects on OS cell proliferation and apoptosis. Furthermore, this study identified EMMPRIN as a target of PG action, through a pathway involving downregulation of EMMPRIN by modulating SP1 through the ROS/miR-27/ZBTB10 axis.

Materials and methods

Cell culture

Human osteosarcoma MG-63 cells were obtained from the Beijing Institute for Cancer Research (Beijing, China), and cultured in DMEM (Invitrogen) supplemented with heat-inactivated 10% fetal bovine serum (Invitrogen) at 37°C in a humidified environment containing 5 % CO2.

Cell proliferation assay

Cell proliferation was assessed using a WST-8 Cell Counting Kit-8 (CCK-8) (Beyotime, Nantong, China). Briefly, 3×103 cells resuspended in 100 μl DMEM containing 10% fetal bovine serum were seeded in 96-well plates, and incubated at various times. Then, 10 μl CCK-8 solution was added to each well for 1 hour at 37°C. Absorbance at 450 nm was measured on an ELX-800 spectrometer reader (BioTek Instru- ments, Winooski, USA).

Cell apoptosis assessment

The proapoptotic effect of PG was determined by flow cytometry (FITC Annexin V apoptosis kit,

BD Pharmingen, NJ, USA). Briefly, cells were rinsed with ice-cold PBS buffer and resuspend-ed in binding buffer at a final density of 1×106 cells/ml. Then, cells were stained with Annexin V-FITC and propidium iodide (PI) for 15 minutes in the dark and analyzed on a flow-cytometer (Beckman Coulter Inc., FL, USA). Annexin V-FITC positive cells were considered to be apoptotic, while those negative for FITC were regarded as living cells.

Apoptosis detection by morphological changes using Hoechst staining

Apoptotic cells were confirmed by Hoechst 33258 staining. Apoptosis was indicated by the presence of condensed or fragmented nuclei which bind Hoechst 33258 with high affinity. MG-63 cells were treated with various PG con-centrations for 48 h, washed with PBS, and fixed with pre-cooled methanol at 500 μl/well for 10 min. Afterwards, cells were stained with 1 μM Hoechst 33258 (Sigma-Aldrich, MO, USA) for 10 min and analyzed on a Leica fluores-cence microscope. Two hundred cells in three randomly selected fields were counted and scored for the incidence of apoptotic chromatin.

Caspase-3 and caspase-9 activity quantitation

To assess the activities of caspase-3 and cas-pase-9, cytosolic proteins were extracted from cells using a hypotonic cell lysis buffer. Then, cytosolic extracts containing 30 µg of protein were analyzed using a colorimetric assay kit specific for caspase-3 and caspase-9 (Ray Biotech, Guangzhou, China).

Mitochondrial membrane potential (MMP) as-sessment

Changes in MMP were examined using the fluo-rochrome dye JC-1 following a standard proto-col. Briefly, MG-63 cells were challenged with PG for 48 hours before incubation with JC-1. The cells were then rinsed with PBS to remove excess dye before quantitation of fluorescence signals by flow cytometry.

Determination of miRNA and mRNA expres-sion levels

Gene expression was assessed by quantitative real time PCR (qPCR) using gene-specific prim-ers as described previously [23]. In brief, total RNA was extracted using a commercial kit

PG induce apoptosis in osteosarcoma cell

1333 Am J Cancer Res 2016;6(6):1331-1344

(RNeasy Mini kit, Qiagen, Dusseldorf, Germany). For miRNA expression analysis, 40 ng of cDNA, obtained by reverse-transcription, was used as a template for PCR [23]. For mRNA quantita-tion, primers for EMMPRIN and Sp1 were syn-thesized based on published sequences [24]. First-strand cDNA was synthesized from 1 μg RNA using the Reverse Transcription System (Takara, Dalian, China). The resulting cDNA (2 μg) was subjected to PCR amplification. The PCR reactions contained SYBR GREEN Master Mix (Solarbio Co., Beijing, China), forward and reverse primers, and 10 ng of template cDNA. PCR was carried out for 5 minutes at 95°C, fol-lowed by 40 cycles of 95°C for 30 seconds, 60°C for 30 s, and 72°C for 30 seconds. Gene expression was analyzed with U6 or GAPDH as internal controls.

Plasmid construction and cell transfection

To assess the role of EMMMPRIN/Sp1 in PG- induced apoptosis in MG-63 cells, EMMMPRIN/Sp1 was overexpressed as previously described [25]. Briefly, full-length cDNA was obtained by reverse transcription, amplified with specific EMMMPRIN/Sp1 primers, and inserted into the pEGFP-N1 vector (Takara Biomedical Techno- logy Co., Ltd., Beijing, China). The resulting plas-mid was named pEGFP-N1-EMMMPRIN/Sp1, and cloned into MG-63 cells to induce EMM- MPRIN/Sp1 expression. MG-63 cells were transfected with empty pEGFP-N1 were used as controls. 48 hours after transfection, G418 was used to select stable clones.

EMMPRIN/Sp1 gene silencing

shRNA oligos for EMMPRIN/Sp1 gene knock-down were designed as previously described [26]. Two different shRNA sequences and a scramble control sequence were subcloned into the plasmid vector pGCsi-H1 following the manufacturer’s instructions, and designated p-shRNA1, p-shRNA2 and p-shRNA-control, respectively. MG-63 cells in logarithmic growth phase were seeded in 6-well plates at a density of 3×105 cells per well, incubated overnight, and transfected with p-shRNA1, p-shRNA2 and p-shRNA-control, respectively, using Lipofecta- mine 2000 (Invitrogen, CA, USA) according to the manufacturer’s protocol. Transfected cells were incubated for 48 hours and transfection efficiency was examined by florescence micros-copy. Stably transfected clones, designated

MG-63/shRNA1, MG-63/shRNA2 and MG-63/shRNA-control, respectively, were verified by quantitative RT-PCR and Western blot. Stable cell lines with higher suppression efficiency were used in subsequent experiments.

Quantitation of reactive oxygen species (ROS) levels

The intracellular changes in ROS generation were assessed by staining the cells with 2,7 dichlorodihydrofluorescein-diacetate (DCFH- DA). Briefly, MG-63 cells were treated with dif-ferent PG concentrations, and further incubat-ed with 20 µM DCFH-DA at 37°C for 30 min-utes. Finally, cells were analyzed by flow cytometry.

miR-27a transfection

The lentiviral constructs miR-27a mimic and miR-con, as well as the miR-27a inhibitor were obtained from Qiagen (Dusseldorf, Germany). MG-63 cells were transfected with miR-27a mimic, miR-27a inhibitor or control miRNA, respectively, using Lipofectamine-2000, accor- ding to the manufacturer’s instructions. Transfection efficiency was confirmed by qPCR.

Western blot

After treatment, cells were lysed with ice-cold lysis buffer (Beyotime, Shanghai, China) for 30 minute on ice. Cell lysates was then collected by centrifugation at 4°C and 12,000 rpm for 5 minute. Equal amounts (50 μg) of lysate pro-teins were separated on 10% SDS-PAGE gels, and transferred onto PVDF membranes (Millipore, MA, USA). Proteins were probed with specific primary antibodies following standard protocols. After washing with TBST, secondary antibodies were added for 2 hours. The blots were washed with TBST before signal detection with chemiluminescence substrate (KPL., Guildford, UK). The BandScan software (Glyko, Novato, CA) was used to quantify protein band intensities, with β-actin included as a loading control.

In vivo anti-tumor activity assay

Four-week-old male nude mice (BALB/c, nu/nu; SPF laboratory animal center of Soochow University, Suzhou, China) were housed under

PG induce apoptosis in osteosarcoma cell

1334 Am J Cancer Res 2016;6(6):1331-1344

Figure 1. PG reduces cell viability and induces apoptosis in MG-63 cells. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml, ^P<0.05 vs. PG, ^^P<0.01 vs. PG. MG-63 cells were incubated with PG at indicated dosages for 48 hours unless otherwise stated.

PG induce apoptosis in osteosarcoma cell

1335 Am J Cancer Res 2016;6(6):1331-1344

pathogen-free conditions. All animal experi-ments were approved by the Animal Ethics Committee of Soochow University. A total of 1.5 × 106 MG-63 cells resuspended in 50% Matrigel were subcutaneously injected into the animals. Tumor volumes were calculated as ½ a2b (a and b are the short and long tumor axes, respectively). After euthanasia, the tumors were harvested for RNA preparation and EMMPRIN, Sp1 and miR-27a gene expression assessment by qRT-PCR.

Statistical analysis

All experiments were performed three times in triplicate unless otherwise stated. Data are mean ± SD (standard deviation), and were com-pared by one-way ANOVA using the SPSS 13.0 software. P<0.05 was considered statistically significant.

Results

PG decreases MG-63 cell viability in a time- and concentration-dependent manner

To assess the effect of PG on cell viability, MG-63 cells were exposed to various concen-trations (0, 20, 50, 100 μg/ml) of PG for 24, 48 or 72 hours. As shown in Figure 1A, PG reduced MG-63 cell viability in a time- and concentra-tion-dependent manner. PG at 50 and 100 μg/ml could significantly suppress MG-63 cell pro-liferation after 24 h of treatment. When treat-ment was prolonged to 48 hours, all PG dosag-es caused a significant loss of cell viability. As expected, cell proliferation was further decreased after treatment with PG for 72 hours. To exclude the time-caused loss of cell viability, 48 hours was selected for subsequent experiments.

PG induces apoptosis via the intrinsic mito-chondrial pathway

Next, we investigated the proapoptotic effect of PG at different concentrations. As shown in Figure 1B, PG induced apoptosis in MG-63 cells in a dose-dependent manner. PG treat-ment at all three tested dosages induced a sig-nificant increase in MG-63 cell apoptosis rate compared with the vehicle control. The pro-apoptotic effect of PG was also confirmed by Hoechst staining (Figure 1C). Since apoptotic cell death is featured by cleavage of caspase-3

and PARP, activation of caspase-3 and PARP was examined to confirm apoptosis. Our results showed that PG increased caspase-3 activa-tion in a dose-dependent manner (Figure 1D). In addition, Western blot results showed that PG treatment was associated with activation of both caspase-3 and PARP (Figure 1E). Since apoptosis in tumor cells can occur via caspase-independent or caspase-dependent pathway [27], a caspase inhibitor (Z-VAD-FMK) was uti-lized to fully explore the underlying mechanism of the proapoptotic effect of PG. As shown in Figure 1F, Z-VAD-FMK treatment significantly abolished the proapoptotic effect of PG, sug-gesting that PG treatment induced apoptosis in MG-63 cells by activating the caspase cascade.

Apoptosis may occur through either intrinsic or extrinsic pathway, depending on mitochondrial involvement [28]. Mitochondria-mediated (or intrinsic) apoptosis is evidenced by caspase-9 activation, decreased MMP levels, and cyto-chrome C translocation from mitochondria to the cytosol. Therefore, to further explore the role of mitochondria in PG-induced apoptosis, changes in caspase-9 activity, MMP levels and cytosolic cytochrome C amounts were exam-ined. Our results showed that PG treatment led to dose-dependent loss of MMP, caspase-9 activation, and increased cytosolic cytochrome C amounts compared with cells treated with the vehicle (Figure 2A-C). To confirm these find-ings, activation of caspase-8, an indicator of extrinsic apoptosis, was also examined. As shown in Figure 2D, PG caused no significant change in activated caspase-8 levels, further demonstrating that PG induced apoptosis in MG-63 cells via the mitochondrial pathway.

PG induces apoptosis in MG-63 cells by modu-lating EMMPRIN

The involvement of EMMPRIN in tumor cell apoptosis was previously reported [15, 29]. In addition, a recent study found that PG prevents hypoxia-induced epithelial-mesenchymal tran-sition in colorectal cancer HCT116 cells by modulating EMMPRIN [30]. Therefore, we assessed whether PG induced apoptosis by modulating EMMPRIN. As shown in Figure 3A and 3B, PG treatment reduced EMMPRIN expression at both mRNA and protein levels. To further explore the role of EMMPRIN in

PG induce apoptosis in osteosarcoma cell

1336 Am J Cancer Res 2016;6(6):1331-1344

Figure 2. PG induces apoptosis via the intrinsic mitochondrial pathway. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml. MG-63 cells were incubated with PG at indicated dosages for 48 hours unless otherwise started.

PG induce apoptosis in osteosarcoma cell

1337 Am J Cancer Res 2016;6(6):1331-1344

PG-induced apoptosis, MG-63 cells were trans-fected with an EMMPRIN overexpression plas-mid. Interestingly, PG-induced apoptosis was significantly abolished by EMMPRIN overex-pression (Figure 3C). In contrast, EMMPRIN knockdown by shRNA led to a significantly increased apoptotic population (Figure 3C), supporting the role of EMMPRIN in apoptosis modulation. Meanwhile, EMMPRIN overexpres-sion in MG-63 cells significantly blocked PG-induced activation of caspase-3 (Figure 3D). Together, our results suggested that PG induced apoptosis in MG-63 cells, at least part-ly, by modulating EMMPRIN.

PG modulates EMMPRIN expression by down-regulating Sp1

The transcription factor Sp1 has been shown to regulate EMMPRIN expression in human malig-nancies [31]. Here, we examined the role of Sp1 in the regulatory effect of PG on EMMPRIN. As shown in Figure 4A and 4B, a dose-depen-dent decrease in Sp1 expression was observed in MG-63 cells treated with PG. Next, Sp1 expression was artificially manipulated to assess its role in EMMPRIN regulation. Our results showed that Sp1 knockdown led to sig-nificantly reduced EMMPRIN levels (Figure 5C and 5D). In addition, the suppressing effect of PG on EMMPRIN expression was significantly

attenuated by Sp1 overexpression in MG-63 cells (Figure 4C and 4D). Moreover, Sp1 overex-pression in Mg-63 cells significantly attenuated the proapoptotic effect of PG in MG-63 cells (Figure 4E). Taken together, the above findings confirmed the hypothesis that PG induced apoptosis by modulating EMMPRIN via Sp1.

PG modulates Sp1 expression by inducing the Sp repressor ZBTB10 via ROS/miR-27a signal-ing

A recent study highlighted that natural prod-ucts could mediate Sp1 downregulation in tumor cells by generating ROS [32]. In addition, a PG related chemical, PG, induces ROS in tumor cells [33]. Therefore, we explored wheth-er ROS induction was involved in PG-mediated downregulation of Sp1. As shown in Figure 5A, PG caused a significant accumulation of intra-cellular ROS. Then, the ROS activator CoCl2 and inhibitor NAC were used to further explore the role of ROS in reducing PG effect on Sp1 and EMMPRIN expression. As shown in Figure 5B and 5C, increasing ROS levels by CoCl2 treat-ment resulted in significantly increased Sp1 and EMMPRIN levels both at the mRNA and protein levels, similar to PG. Meanwhile, treat-ment with both PG and NAC almost completely abrogated the PG-induced repression on Sp1 and EMMPRIN expression. The above results

Figure 3. Modulation of EMMPRIN by PG is involved in the pro-apoptotic effect of PG. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml, ^^P<0.01 vs. PG. MG-63 cells were incubated with PG at 100 μg/ml for 48 hours un-less otherwise stated.

PG induce apoptosis in osteosarcoma cell

1338 Am J Cancer Res 2016;6(6):1331-1344

indicated that PG downregulated SP1 and EMMPRIN in a ROS-dependent manner. Moreover, our results also demonstrated that ROS generation was involved in PG-induced MG-63 cell apoptosis, since NAC significantly abolished the proapoptotic effect of PG.

Previous studies showed that ROS-dependent downregulation of Sp1, Sp3, and Sp4 in human malignancies is due to reduced miR-27a and/or miR-20a/miR-17 levels, resulting in the induc-tion of miR-regulated “Sp repressors” ZBTB10 and ZBTB4, respectively [32], which competi-tively bind GC-rich Sp binding sites in the pro-moters of Sp1, Sp3, Sp4, and Sp-regulated genes, causing decreased transactivation [34, 35]. Therefore, the effects of PG on miR-27a/ZBTB10 and miR-20a or miR-17/ZBTB4 axis were assessed to explore the mechanism by which PG modulated Sp1 expression. As shown in Figure 6A-C, PG treatment did not signifi-cantly alter ZBTB4 levels and miR-20a or miR-17 expression. In contrast, PG treatment

induced a remarkable increase in ZBTB10 expression as well as significantly decreased miR-27a cell levels (Figure 6A-C). Regulation of SP1 was confirmed by using miR-27a mimics that decreased ZBTB10 expression, abrogating the modulating effect of PG on SP1 (Figure 6D and 6E). In agreement, the suppressive effect of PG on EMMPRIN levels was significantly reduced by transfection with miR-27a mimics (Figure 6D and 6E). Collectively, our results indicated that PG disrupted the miR-27a/ZBTB10 axis, resulting in repressed SP1 and subsequent downregulation of EMMPRIN. As expected, increasing levels of miR-27a using miR-27a mimics could alter PG-induced apop-tosis, indicating miR-27a involvement in the proapoptotic effect of PG (Figure 6F).

In vivo anticancer effect of PG

Next, we sought to assess the in vivo effect of PG. Nude mice were injected with KB cells and tumors were allowed to grow to 100 mm3. Then,

Figure 4. PG modulates EMMPRIN expression by regulation Sp1. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml, ^^P<0.01 vs. PG. MG-63 cells were incubated with PG at 100 μg/ml for 48 hours unless otherwise stated.

PG induce apoptosis in osteosarcoma cell

1339 Am J Cancer Res 2016;6(6):1331-1344

the animals were treated with different doses of PG, including 100 mg, 50 mg, and 10 mg/kg/day, respectively. Treatment efficacy was evaluated by measuring tumor volumes. As shown in Figure 7A, PG inhibited tumor growth dose-dependently, and this anti-tumor effect was significant at 50 and 100 mg/kg/day. Moreover, PCR data showed that PG caused a marked decrease in EMMPRIN mRNA levels, suggesting in vivo PG effectiveness in inhibiting tumor growth was mediated by EMMPRIN downregulation (Figure 7B). Moreover, the inhibitory effect of PG on tumor growth was associated with significantly decreased Sp1 and miR-27a expression levels (Figure 7C), fur-ther supporting our in vitro findings that apop-tosis induced by PG was associated with EMMPRIN downregulation via miR-27a modu- lation.

Discussions

Apoptosis, also called type I programmed cell death, is associated with the activation of cata-bolic enzymes, eventually resulting in nuclear chromatin condensation, nuclear fragmenta-tion, and formation of distinct apoptotic bodies

[36]. Mounting evidence suggests that apopto-sis induction in tumor cells is a major protective mechanism against the development and pro-gression of cancer [37]. Our results revealed that PG exhibited dose- and time-dependent anti-proliferation effects on OS cells by promot-ing apoptosis, suggesting the potential of PG as a chemotherapeutic agent in the treatment of OS. Apoptosis is mainly mediated by the death receptor-triggered extrinsic pathway and mito-chondrial-initiated intrinsic pathway, featured by activation of caspase-8 and caspase-9, respectively [38, 39]. In the present study, the observed PG-mediated caspase-9 and cas-pase-3 activation as well as cytochrome C release into cytosol and loss of MMP suggest that PG induces apoptosis mainly via the mito-chondrial-mediated caspase pathway.

EMMPRIN is highly expressed on the surface of various malignant tumor cells, and mainly func-tions as a cellular adhesion molecule that induces secretion of matrix metalloproteinases (MMP; mainly MMP-1, MMP-2 and MMP-9), thus promoting invasion and metastasis [40, 41]. Besides its function in invasion and metas-tasis [41], the role of EMMPRIN in apoptosis

Figure 5. PG regulates Sp1 by generating ROS. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml, ^^P<0.01 vs. PG. MG-63 cells were incubated with PG at 100 μg/ml for 48 hours unless otherwise stated.

PG induce apoptosis in osteosarcoma cell

1340 Am J Cancer Res 2016;6(6):1331-1344

was demonstrated by RNA interference show-ing that EMMPRIN knockdown induces apopto-sis in both solid tumors and hematological malignancies [26, 42, 43]. In addition, recent studies proposed that natural products exert apoptosis-inducing effects by downregulating EMMPRIN [29, 44]. In this study, the proapop-totic effect of PG in MG-63 cells was mediated, at least partly, by EMMPRIN regulation, corrob-orating recent studies and confirming the role of EMMPRIN in OS cell apoptosis.

Gene regulation by transcription factors is criti-cal to many biological processes and carcino-genesis [45]. One of the first transcription fac-tors identified in mammalian cells is Sp1 [45], a member of the zinc-finger Sp family of proteins that includes the Kruppel-like factor (KLF) fam-ily [46]. Sp1 is expressed ubiquitously in vari-

ous mammalian cells and implicated in the transcription of many genes that contain GC boxes in their promoters [47], particularly housekeeping genes and those involved in cell growth and development. Sp1 is overexpressed in a number of human malignancies, and regu-lates many aspects of cancer biology by regu-lating pro-oncogenic genes important for cell growth (cyclin D1, EGFR, and c-Met) [48], sur-vival (bcl-2 and survivin) [48], angiogenesis (VEGF and VEGF receptors) [49], and metasta-sis [50, 51]. In a previous study, Kong et al reported that differential Sp1 expression and activity directly regulate EMMPRIN levels in human lung cancer [31]. Ke et al also found that Sp1, combined with HIF-1α, induces EMMPRIN expression in hypoxia [52]. As shown above, PG regulated EMMPRIN expression by Sp1 repression; in addition, similar effects were

Figure 6. PG exerts ROS-dependent regulatory effect on Sp1 by modulating miR-27a/ZBTB10. *P<0.05 vs. PG at 0 μg/ml, **P<0.01 vs. PG at 0 μg/ml, ^^P<0.01 vs. PG. MG-63 cells were incubated with PG at 100 μg/ml for 48 hours unless otherwise stated.

PG induce apoptosis in osteosarcoma cell

1341 Am J Cancer Res 2016;6(6):1331-1344

obtained by artificially manipulating Sp1 amounts in MG-63 cells, corroborating previ-ous findings demonstrating the regulatory role of Sp1 in EMMPRIN expression. Furthermore, our findings highlight the potential of PG to interfere with various neoplastic activities through Sp1 repression, thus regulating Sp1-targeting genes.

In normal cells, reactive oxygen species (ROS), including radicals (superoxide, nitric oxide, and hydroxyl radicals) and non-radical species (hydrogen peroxide, ozone, and peroxynitrates) are generated as by-products of cellular metab-olism, and in cellular redox balance with bio-chemical antioxidants [53]. In cancer cells, a modest increase in ROS can enhance cell pro-liferation, survival, and drug resistance; how-ever, further increase of ROS that cannot be

attenuated by intracellular redox systems can induce apoptosis and cell cycle arrest [53]. Given that ROS levels are higher in cancer than in non-cancer cells, ROS-inducing drugs can selectively kill cancer cells without causing overt toxicity to normal cells [53]. Therefore, increasing ROS production is considered an important strategy for cancer therapies. In the context of OS, a number of natural products have been found to induce apoptosis by gener-ating excessive cellular ROS and triggering vari-ous signaling pathways, including JNK [54], p38 MAPK [55], and JAK2/STAT3 [56] path-ways. In this study, PG induced ROS and decreased miR-27a levels in MG-63 cells, which resulted in the upregulation of the Sp-repressor ZBTB10; this resulted in SP1 downregulation and the subsequent repression of EMMPRIN. On the other hand, it has been found that ROS-

Figure 7. In vivo anti-tumor effect of PG is associated with downregulation of EMMPRIN, Sp1 and miR-27a as well as elevated expression of ZBTB10. *P<0.05 vs. Vehicle, **P<0.01 vs. Vehicle.

PG induce apoptosis in osteosarcoma cell

1342 Am J Cancer Res 2016;6(6):1331-1344

inducing agents could exert pro-autophagic effects on OS cells [54, 57, 58]. Therefore, by generating ROS, PG might also kill OS cells by inducing autophagy.

In conclusion, our results showed that PG induces apoptosis in OS cells, at least in part, through a pathway involving EMMPRIN down-regulation, by modulating SP1 via the ROS/miR-27/ZBTB10 axis, highlighting the potential of PG as an anticancer agent. However, further studies, including clinical trials, are needed to fully evaluate PG as a novel therapeutic in can-cer prevention and treatment.

Disclosure of conflict of interest

None.

Address correspondence to: Huilin Yang, Depart- ment of Orthopaedic Surgery, The First Affiliated Hospital of Soochow University, No 188, Shizi Street, Suzhou 215006, China. E-mail: hcooper@126.com

References

[1] Rabinowicz R, Barchana M, Liphshiz I, Futerman B, Linn S and Weyl-Ben-Arush M. Cancer incidence and survival among children and adolescents in Israel during the years 1998 to 2007. J Pediatr Hematol Oncol 2012; 34: 421-429.

[2] Mirabello L, Troisi RJ and Savage SA. International osteosarcoma incidence pat-terns in children and adolescents, middle ages and elderly persons. Int J Cancer 2009; 125: 229-234.

[3] Carroll KL, Yandow SM, Ward K and Carey JC. Clinical correlation to genetic variations of he-reditary multiple exostosis. J Pediatr Orthop 1999; 19: 785-791.

[4] Kager L, Zoubek A, Kastner U, Kempf-Bielack B, Potratz J, Kotz R, Exner GU, Franzius C, Lang S, Maas R, Jurgens H, Gadner H, Bielack S; Cooperative Osteosarcoma Study Group. Skip metastases in osteosarcoma: experience of the Cooperative Osteosarcoma Study Group. J Clin Oncol 2006; 24: 1535-1541.

[5] Leary SE, Wozniak AW, Billups CA, Wu J, McPherson V, Neel MD, Rao BN and Daw NC. Survival of pediatric patients after relapsed os-teosarcoma: the St. Jude Children’s Research Hospital experience. Cancer 2013; 119: 2645-2653.

[6] Jia L, Ma S, Hou X, Wang X, Qased AB, Sun X, Liang N, Li H, Yi H, Kong D, Liu X and Fan F. The synergistic effects of traditional Chinese herbs and radiotherapy for cancer treatment. Oncol Lett 2013; 5: 1439-1447.

[7] Shi RX, Ong CN and Shen HM. Luteolin sensi-tizes tumor necrosis factor-alpha-induced apoptosis in human tumor cells. Oncogene 2004; 23: 7712-7721.

[8] Lin SY, Liu JD, Chang HC, Yeh SD, Lin CH and Lee WS. Magnolol suppresses proliferation of cultured human colon and liver cancer cells by inhibiting DNA synthesis and activating apop-tosis. J Cell Biochem 2002; 84: 532-544.

[9] Ranuncolo SM, Armanasco E, Cresta C, Bal De Kier Joffe E and Puricelli L. Plasma MMP-9 (92 kDa-MMP) activity is useful in the follow-up and in the assessment of prognosis in breast cancer patients. Int J Cancer 2003; 106: 745-751.

[10] Zhu H, Evans B, O’Neill P, Ren X, Xu Z, Hait WN and Yang JM. A role for p53 in the regulation of extracellular matrix metalloproteinase inducer in human cancer cells. Cancer Biol Ther 2009; 8: 1722-1728.

[11] Zheng HC, Takahashi H, Murai Y, Cui ZG, Nomoto K, Miwa S, Tsuneyama K and Takano Y. Upregulated EMMPRIN/CD147 might con-tribute to growth and angiogenesis of gastric carcinoma: a good marker for local invasion and prognosis. Br J Cancer 2006; 95: 1371-1378.

[12] Tang Y, Nakada MT, Kesavan P, McCabe F, Millar H, Rafferty P, Bugelski P and Yan L. Extracellular matrix metalloproteinase inducer stimulates tumor angiogenesis by elevating vascular endothelial cell growth factor and ma-trix metalloproteinases. Cancer Res 2005; 65: 3193-3199.

[13] Duivenvoorden WC, Hirte HW and Singh G. Transforming growth factor beta1 acts as an inducer of matrix metalloproteinase expres-sion and activity in human bone-metastasizing cancer cells. Clin Exp Metastasis 1999; 17: 27-34.

[14] Quemener C, Gabison EE, Naimi B, Lescaille G, Bougatef F, Podgorniak MP, Labarchede G, Lebbe C, Calvo F, Menashi S and Mourah S. Extracellular matrix metalloproteinase inducer up-regulates the urokinase-type plasminogen activator system promoting tumor cell inva-sion. Cancer Res 2007; 67: 9-15.

[15] Liang H, Zhan HJ, Wang BG, Pan Y and Hao XS. Change in expression of apoptosis genes after hyperthermia, chemotherapy and radiotherapy in human colon cancer transplanted into nude mice. World J Gastroenterol 2007; 13: 4365-4371.

[16] Misra S, Ghatak S, Zoltan-Jones A and Toole BP. Regulation of multidrug resistance in can-cer cells by hyaluronan. J Biol Chem 2003; 278: 25285-25288.

[17] Lu Q, Lv G, Kim A, Ha JM and Kim S. Expression and clinical significance of extracellular matrix metalloproteinase inducer, EMMPRIN/CD147,

PG induce apoptosis in osteosarcoma cell

1343 Am J Cancer Res 2016;6(6):1331-1344

in human osteosarcoma. Oncol Lett 2013; 5: 201-207.

[18] Zhou Q, Zhu Y, Deng Z, Long H, Zhang S and Chen X. VEGF and EMMPRIN expression cor-relates with survival of patients with osteosar-coma. Surg Oncol 2011; 20: 13-19.

[19] Jiang L, Zhang S and Xuan L. Oxanthrone C-glycosides and epoxynaphthoquinol from the roots of Rumex japonicus. Phytochemistry 2007; 68: 2444-2449.

[20] Zhou X, Xuan L and Zhang S. [Study on the chemical constituents from Rumex japonicus Houtt]. Zhong Yao Cai 2005; 28: 104-105.

[21] Belkin M and Fitzgerald DB. Tumor-damaging capacity of plant materials. I. Plants used as cathartics. J Natl Cancer Inst 1952; 13: 139-155.

[22] Xie QC and Yang YP. Anti-proliferative of phy-scion 8-O-beta-glucopyranoside isolated from Rumex japonicus Houtt. on A549 cell lines via inducing apoptosis and cell cycle arrest. BMC Complement Altern Med 2014; 14: 377.

[23] Yang CH, Yue J, Fan M and Pfeffer LM. IFN in-duces miR-21 through a signal transducer and activator of transcription 3-dependent path-way as a suppressive negative feedback on IFN-induced apoptosis. Cancer Res 2010; 70: 8108-8116.

[24] Sato M, Nakai Y, Nakata W, Yoshida T, Hatano K, Kawashima A, Fujita K, Uemura M, Takayama H and Nonomura N. EMMPRIN pro-motes angiogenesis, proliferation, invasion and resistance to sunitinib in renal cell carci-noma, and its level predicts patient outcome. PLoS One 2013; 8: e74313.

[25] Hou Q, Tang X, Liu H, Tang J, Yang Y, Jing X, Xiao Q, Wang W, Gou X and Wang Z. Berberine in-duces cell death in human hepatoma cells in vitro by downregulating CD147. Cancer Sci 2011; 102: 1287-1292.

[26] Gao H, Jiang Q, Han Y, Peng J and Wang C. shRNA-mediated EMMPRIN silencing inhibits human leukemic monocyte lymphoma U937 cell proliferation and increases chemosensitiv-ity to adriamycin. Cell Biochem Biophys 2015; 71: 827-835.

[27] Zhou H, Xu M, Gao Y, Deng Z, Cao H, Zhang W, Wang Q, Zhang B, Song G, Zhan Y and Hu T. Matrine induces caspase-independent pro-gram cell death in hepatocellular carcinoma through bid-mediated nuclear translocation of apoptosis inducing factor. Mol Cancer 2014; 13: 59.

[28] Elmore S. Apoptosis: a review of programmed cell death. Toxicol Pathol 2007; 35: 495-516.

[29] Chen X, Gao H, Han Y, Ye J, Xie J and Wang C. Physcion induces mitochondria-driven apopto-sis in colorectal cancer cells via downregulat-ing EMMPRIN. Eur J Pharmacol 2015; 764: 124-133.

[30] Ding Z, Xu F, Tang J, Li G, Jiang P, Tang Z and Wu H. Physcion 8-O-beta-glucopyranoside pre-vents hypoxia-induced epithelial-mesenchymal transition in colorectal cancer HCT116 cells by modulating EMMPRIN. Neoplasma 2016; 63: 351-61.

[31] Kong LM, Liao CG, Fei F, Guo X, Xing JL and Chen ZN. Transcription factor Sp1 regulates expression of cancer-associated molecule CD147 in human lung cancer. Cancer Sci 2010; 101: 1463-1470.

[32] Gandhy SU, Kim K, Larsen L, Rosengren RJ and Safe S. Curcumin and synthetic analogs induce reactive oxygen species and decreases specificity protein (Sp) transcription factors by targeting microRNAs. BMC Cancer 2012; 12: 564.

[33] Wijesekara I, Zhang C, Van Ta Q, Vo TS, Li YX and Kim SK. Physcion from marine-derived fungus Microsporum sp. induces apoptosis in human cervical carcinoma HeLa cells. Microbiol Res 2014; 169: 255-261.

[34] Mertens-Talcott SU, Chintharlapalli S, Li X and Safe S. The oncogenic microRNA-27a targets genes that regulate specificity protein tran-scription factors and the G2-M checkpoint in MDA-MB-231 breast cancer cells. Cancer Res 2007; 67: 11001-11011.

[35] Kim K, Chadalapaka G, Lee SO, Yamada D, Sastre-Garau X, Defossez PA, Park YY, Lee JS and Safe S. Identification of oncogenic microR-NA-17-92/ZBTB4/specificity protein axis in breast cancer. Oncogene 2012; 31: 1034-1044.

[36] Maiuri MC, Zalckvar E, Kimchi A and Kroemer G. Self-eating and self-killing: crosstalk be-tween autophagy and apoptosis. Nat Rev Mol Cell Biol 2007; 8: 741-752.

[37] Ghobrial IM, Witzig TE and Adjei AA. Targeting apoptosis pathways in cancer therapy. CA Cancer J Clin 2005; 55: 178-194.

[38] Dias N and Bailly C. Drugs targeting mitochon-drial functions to control tumor cell growth. Biochem Pharmacol 2005; 70: 1-12.

[39] Schulze-Osthoff K, Ferrari D, Los M, Wesselborg S and Peter ME. Apoptosis signaling by death receptors. Eur J Biochem 1998; 254: 439-459.

[40] Kuang YH, Chen X, Su J, Wu LS, Liao LQ, Li D, Chen ZS and Kanekura T. RNA interference tar-geting the CD147 induces apoptosis of multi-drug resistant cancer cells related to XIAP de-pletion. Cancer Lett 2009; 276: 189-195.

[41] Sun J and Hemler ME. Regulation of MMP-1 and MMP-2 production through CD147/extra-cellular matrix metalloproteinase inducer inter-actions. Cancer Res 2001; 61: 2276-2281.

[42] Xu X, Liu S, Lei B, Li W, Lin N, Sheng W, Huang A and Shen H. Expression of HAb18G in non-small lung cancer and characterization of acti-

PG induce apoptosis in osteosarcoma cell

1344 Am J Cancer Res 2016;6(6):1331-1344

vation, migration, proliferation, and apoptosis in A549 cells following siRNA-induced down-regulation of HAb18G. Mol Cell Biochem 2013; 383: 1-11.

[43] Zhao Y, Chen S, Gou WF, Niu ZF, Zhao S, Xiao LJ, Takano Y and Zheng HC. The role of EMMPRIN expression in ovarian epithelial car-cinomas. Cell Cycle 2013; 12: 2899-2913.

[44] Lv C, Kong H, Dong G, Liu L, Tong K, Sun H, Chen B, Zhang C and Zhou M. Antitumor effi-cacy of alpha-solanine against pancreatic can-cer in vitro and in vivo. PLoS One 2014; 9: e87868.

[45] Dynan WS and Tjian R. Isolation of transcrip-tion factors that discriminate between differ-ent promoters recognized by RNA polymerase II. Cell 1983; 32: 669-680.

[46] Govindaraju S, Lee BS. Krüppel-like factor 8 is a stress-responsive transcription factor that regulates expression of HuR. Cell Physiol Biochem 2014; 34: 519-32.

[47] Suske G. The Sp-family of transcription factors. Gene 1999; 238: 291-300.

[48] Jutooru I, Guthrie AS, Chadalapaka G, Pathi S, Kim K, Burghardt R, Jin UH and Safe S. Mechanism of action of phenethylisothiocya-nate and other reactive oxygen species-induc-ing anticancer agents. Mol Cell Biol 2014; 34: 2382-2395.

[49] Chintharlapalli S, Papineni S, Ramaiah SK and Safe S. Betulinic acid inhibits prostate cancer growth through inhibition of specificity protein transcription factors. Cancer Res 2007; 67: 2816-2823.

[50] Ibanez-Tallon I, Ferrai C, Longobardi E, Facetti I, Blasi F and Crippa MP. Binding of Sp1 to the proximal promoter links constitutive expres-sion of the human uPA gene and invasive po-tential of PC3 cells. Blood 2002; 100: 3325-3332.

[51] Price SJ, Greaves DR and Watkins H. Identification of novel, functional genetic vari-ants in the human matrix metalloproteinase-2 gene: role of Sp1 in allele-specific transcrip-tional regulation. J Biol Chem 2001; 276: 7549-7558.

[52] Ke X, Fei F, Chen Y, Xu L, Zhang Z, Huang Q, Zhang H, Yang H, Chen Z and Xing J. Hypoxia upregulates CD147 through a combined effect of HIF-1alpha and Sp1 to promote glycolysis and tumor progression in epithelial solid tu-mors. Carcinogenesis 2012; 33: 1598-1607.

[53] Trachootham D, Alexandre J and Huang P. Targeting cancer cells by ROS-mediated mech-anisms: a radical therapeutic approach? Nat Rev Drug Discov 2009; 8: 579-591.

[54] Li HY, Zhang J, Sun LL, Li BH, Gao HL, Xie T, Zhang N and Ye ZM. Celastrol induces apopto-sis and autophagy via the ROS/JNK signaling pathway in human osteosarcoma cells: an in vitro and in vivo study. Cell Death Dis 2015; 6: e1604.

[55] Hu KH, Li WX, Sun MY, Zhang SB, Fan CX, Wu Q, Zhu W and Xu X. Cadmium Induced Apoptosis in MG63 Cells by Increasing ROS, Activation of p38 MAPK and Inhibition of ERK 1/2 Pathways. Cell Physiol Biochem 2015; 36: 642-654.

[56] Liu Y, Wang L, Wu Y, Lv C, Li X, Cao X, Yang M, Feng D and Luo Z. Pterostilbene exerts antitu-mor activity against human osteosarcoma cells by inhibiting the JAK2/STAT3 signaling pathway. Toxicology 2013; 304: 120-131.

[57] Niu NK, Wang ZL, Pan ST, Ding HQ, Au GH, He ZX, Zhou ZW, Xiao G, Yang YX, Zhang X, Yang T, Chen XW, Qiu JX and Zhou SF. Pro-apoptotic and pro-autophagic effects of the Aurora ki-nase A inhibitor alisertib (MLN8237) on hu-man osteosarcoma U-2 OS and MG-63 cells through the activation of mitochondria-mediat-ed pathway and inhibition of p38 MAPK/PI3K/Akt/mTOR signaling pathway. Drug Des Devel Ther 2015; 9: 1555-1584.

[58] Liu Y, Zhao L, Ju Y, Li W, Zhang M, Jiao Y, Zhang J, Wang S, Wang Y, Zhao M, Zhang B and Zhao Y. A novel androstenedione derivative induces ROS-mediated autophagy and attenuates drug resistance in osteosarcoma by inhibiting mac-rophage migration inhibitory factor (MIF). Cell Death Dis 2014; 5: e1361.