MOLECULAR CARCINOGENESIS

GOLPH3 Promotes Glioblastoma Cell Migration and Invasion via the mTOR-YB1 Pathway In Vitro Xu Zhang,1,2 Zhijun Ding,3 Jianbing Mo,3 Ben Sang,3 Qiong Shi,1,2 Jinxia Hu,1,2 Shao Xie,1,2 Wenjian Zhan,1, 2 Dong Lu,3 Minglin Yang,3 Wenbin Bian,3 Xiuping Zhou,1,2* and Rutong Yu1,2** 1

Brain Hospital, Affiliated Hospital of Xuzhou Medical College, Xuzhou, Jiangsu, China Insititute of Nervous System Diseases, Xuzhou Medical College, Xuzhou, Jiangsu, China 3 The Graduate School, Xuzhou Medical College, Xuzhou, Jiangsu, China 2

The identification of genes involved in carcinogenesis and tumor progression is of great interest, since these genes might be possible as candidates for new tumor targeted therapy strategies. Our previous study shows that Golgi phosphoprotein 3 (GOLPH3) is involved in glioma cell migration and invasion, the critical characteristics of malignant gliomas. In this study, we explored the mechanism of GOLPH3 affecting cell migration and invasion and found that GOLPH3 promotes glioblastoma (GBM) cell migration and invasion via the mammalian target of rapamycin(mTOR)-Y-box binding protein-1 (YB1) pathway in vitro. Both the protein levels of GOLPH3 and YB1 were up-regulated in human glioma tissues and they exhibited direct correlation with each other. In addition, down-regulation of GOLPH3 inhibited glioma cell migration and invasion, while over-expression of GOLPH3 enhanced them. Meanwhile, GOLPH3 down-regulation led to a significant decrease of YB1 level as well as mTOR activity, both required for glioma cell migration and invasion. On the contrary, YB1 level and mTOR activity increased after GOLPH3 over-expression. YB1 down-regulation or mTOR ATP site inhibitor INK128 treatment inhibited cell migration and invasion, similar to the effect of GOLPH3 down-regulation. Furthermore, over-expression of GOLPH3 induced glioma cell migration and invasion was blocked by INK128 and YB1 down-regulation. Taken together, these results show that GOLPH3 promotes glioblastoma cell migration and invasion via the mTORYB1pathway, indicating that GOLPH3-mTOR-YB1 pathway might be a new therapeutic target for glioma treatment. © 2014 Wiley Periodicals, Inc.

Key words: glioma; GOLPH3; mTOR; YB1; migration; invasion

INTRODUCTION Glioblastoma (GBM) is the most common and malignant primary central nervous system tumor, with a disproportionate level of morbidity and mortality across a wide range of individuals [1]. Despite advances in surgery and adjuvant therapy, the median overall survival time does not exceed 15 months [2–4]. One of the major reasons for the failure of traditional treatments is the highly invasive nature of glioma cells [5]. Golgi phosphoprotein 3 (GOLPH3, also called GPP34, GMx33, and MIDAS) is one of the many proteins in the trans-Golgi matrix involved in anterograde and retrograde Golgi traffic and in interactions with the cytoskeleton to maintain Golgi structure [6]. GOLPH3 is found to bind to phosphatidylinositol 4-phosphate and to play a vital role in the Golgi secretory pathway and protein glycosylation [7–10]. Recently, GOLPH3 is identified as a novel oncogene by genome-wide array-based comparative genome hybridization and tumor tissue microarrays. GOLPH3 promotes cell transformation and tumor growth by constitutively activating mammalian target of rapamycin(mTOR) signaling and conferring increased sensitivity to rapamycin in melanoma ß 2014 WILEY PERIODICALS, INC.

cells [11]. Human tumor cell lines over-expressing GOLPH3 develop tumors faster than controls and are more sensitive to rapamycin [11,12]. Later studies further demonstrate that GOLPH3 plays important roles in the progression of many tumors [13,14]. For

Abbreviations: GOLPH3, Golgi phosphoprotein 3; mTOR, mammalian target of rapamycin; YB1, Y-box binding protein-1; GBM, Glioblastoma. Conflict of interest: The authors declare no conflict of interest. Grant sponsor: National Natural Science Foundation of China; Grant number: 81272777; 81372699; Grant sponsor: Natural Science Foundation of Jiangsu Province; Grant number: BK2011195; Grant sponsor: Program for New Century Excellent Talents in University; Grant number: NCET-10-0181; Grant sponsor: Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). *Correspondence to: Institute of Nervous System Diseases, Xuzhou Medical College, 84 West Huai-hai Road, Xuzhou, Jiangsu, 221002, PR China. **Correspondence to: Brain Hospital, Affiliated Hospital of Xuzhou Medical College, 99 West Huai-hai Road, Xuzhou, Jiangsu, 221002, PR China. Received 6 January 2014; Revised 29 May 2014; Accepted 2 June 2014 DOI 10.1002/mc.22197 Published online in Wiley Online Library (wileyonlinelibrary.com).

2

ZHANG ET AL.

the first time, our previous study finds that GOLPH3 is involved in glioma cell migration and invasion [15]. However, how GOLPH3 regulate cell migration and invasion is largely unknown. Mammalian Y-box binding protein-1 (YB1) is a member of the DNA/RNA-binding family of proteins with an evolutionarily conserved cold-shock domain. Both bacterial and mammalian cold-shock domain proteins are ubiquitously expressed and involved in fundamental processes such as DNA repair, mRNA transcription, splicing, translation and stabilization [16,17]. In other words, it performs both overall and specific regulation of gene expression at different levels [18]. It is reported that YB1 translocation from the cytoplasm to the nucleus stimulates transcription of a number of genes encoding proteins responsible for multiple drug resistance [19]. When involved in DNA reparation in the nucleus, YB1 contributes to cell resistance against ionizing radiation and xenobiotics [20]. Therefore, the nuclear localization of YB1 is considered to be an early marker of multiple drug resistance of cancer cells [19,21]. In addition, previous studies show that the amount of YB1 is especially high in cancer cells and is associated with poor prognosis, disease recurrence, and drug resistance [22–24]. Its elevated level in the cytoplasm may prevent oncogenic transformation of cells caused by activated PI3K/Akt kinase signaling pathway [25]. Furthermore, it is reported that YB1 plays a key role in glioma migration and invasion, which promotes epithelialmesenchymal transition (EMT), a process whereby epithelial cells lose polarity, cell-cell contacts, and cytoskeletal integrity and acquire metastatic ability [26,27]. Although the link between YB1 upregulation and tumor aggressiveness is well established [28,29], the precise mechanism remains largely unknown. Here, we show that GOLPH3 promotes glioblastoma cell migration and invasion via the mTOR-YB1 pathway.

MATERIALS AND METHODS Glioma and Nontumorous Human Brain Tissues Surgically removed human glioma tissue samples and nontumorous brain tissues were obtained from the affiliated hospital of Xuzhou Medical College (Xuzhou, China). Gliomas were graded by the Department of Pathology at the affiliated hospital of Xuzhou Medical College based on the WHO grading system. As controls, human nontumorous brain tissues (mostly from the cortex) were obtained from patients with decompressive surgery after physical injuries to the brain. These specimens were collected from the patients registered at the above-mentioned hospital, and written informed consent was obtained from the patients. This study was approved by the ethics committee of the hospital. Molecular Carcinogenesis

Cell Culture The human glioblastoma cell line U251 and U87 were purchased from Shanghai Institutes for Biological Sciences (SIBS) of Chinese Academy of Sciences (CAS). Cells were cultured in 5% CO2 and 95% humidified atmosphere air at 378C in DMEM/F-12 (Gibco, Carlsbad, CA), supplemented with 10% FBS (Evergreen Biological Engineering Co., Ltd. Hangzhou, China). Cells were split every 3 days to ensure logarithmic growth. Antibodies and Drugs The following antibodies were used: rabbit monoclonal anti-GOLPH3, mouse monoclonal anti-MYC, rabbit polyclonal anti-GOLPH3L, mouse monoclonal anti-4EBP1 and rabbit monoclonal anti-phospho-YB1 (Abcam, Cambridge, MA); mouse monoclonal antib-actin and rapamycin (Millipore, Billerica, MA); mouse monoclonal anti-YB1 (Santa Cruz, CA); rabbit monoclonal anti-phospho-4EBP1, rabbit monoclonal anti-p70S6K, and rabbit monoclonal anti-phosphop70S6K (Cell signaling, Danvers, MA); INK128 (Takeda, Osaka, Japan). Small Interfering RNAs and Plasmids Transfection U251 and U87 cells were transiently transfected with human GOLPH3 siRNA, YB1 siRNA or MYC-GOLPH3 using Lipofectamine 2000 (Invitrogen, Grand Island, NY) according to the manufacturer’s instructions. Some sets of siRNA duplexes (GenePharma Co. Shanghai, China) were listed below: GOLPH3 si571: 50 -GAAUUAGCAUUGAGAGGAATT-30 ; GOLPH3 si635: 50 -CAAGAAAGGUAAUCUGUAATT-30 ; GOLPH3 si792: 50 -GUUAAGAAAUGUACGGGAATT-30 ; YB1 si1015: 50 -GGUUCCCACCUUACUACAUTT-30 . All the transfections were performed three times independently. Immunohistochemistry Paraffin-embedded sections were obtained from the pathology files of the Department of Pathology at the affiliated hospital of Xuzhou Medical College. Nontumorous brain specimens were acquired from patient under-going surgery for internal decompression in cerebral trauma and were reviewed to verify the absence of tumor. Paraffin sections (5 mm thick) were baked at 608C, deparaffinized in xylene and rehydrated in graded ethanol, and then microwaved for antigen retrieval. These sections were treated in 3% hydrogen peroxide for 30 min. Slides were incubated at 48C overnight with anti-GOLPH3 antibody (1:200 dilution) or anti-YB1 antibody (1:100 dilution). The bound antibodies were detected by use of the streptavidin-peroxidase kit (Beijing Zhongshan Golden Bridge Bio). After finishing the steps above, the slides were counterstained with hematoxylin,

GOLPH3-MTOR-YB1 AND GLIOMA CELL MIGRATION AND INVASION

dehydrated with ethanol and xylene, and covered with coverslips. Micrographs were taken and the results were presented as the percentage of the glioma cells with positive staining. Cell Proliferation Assay The effect of GOLPH3,YB1 or mTOR inhibitors on glioma cell proliferation was obtained by detecting the cell viability with the Cell Counting Kit-8 (CCK-8, Beyotime) according to the manufacturer’s instruction. Data were obtained from three independent assays performed in triplicate. Wound Healing Assay Twenty-four hours after transfection, a lesion was created using a plastic pipette tip, and the cells were washed twice with PBS to remove the debris. The monolayer was then maintained in serum-free DMEM and cultured for 24 h. At the designated time, five randomly selected fields at the lesion border were acquired under an inverted microscope (Olympus, IX71, Tokyo, Japan). The number of cells across the wound was counted. Transwell Invasion Assay Cell invasion assay were performed using a transwell system (Corning, NY) according to the manufacturer’s protocol. Filters were precoated with matrigel (BD Biosciences, Franklin Lakes, NJ). Approximately 5  104 cells in serum-free media were added to the top chamber and the bottom chamber was filled with DMEM containing 10% FBS. After 24 h of incubation, the cells on the upper surface were gently removed with a cotton swab and the membrane was fixed in 4% methanol for 30 min and stained with a 0.1% crystal violet solution for 30 min. The cells that migrated to the lower side of the membrane were captured and the cell number was counted. Western Blotting Total proteins were extracted from nontumorous, brain glioma tissues or cells cultured in vitro using the radioimmunoprecipitation assay buffer (150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1 mM sodium orthovanadate, 50 mM NaF, 1 mM phenylmethanesulfonyl fluoride, 1 mM aprotinin, 1 mM leupeptin, 5 mM dithiothreitol, and 10 mM Tris-HCl [pH 7.4]). Protein lysates were concentrated using the BCA Protein Assay Kit (Bio-Rad. Lab. Hercules, CA) and equal amount of protein lysates were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Later, the proteins were transferred to polyvinylidene difluoride membrane of 0.45 mm pore size (Millipore, Billerica, MA), which were blocked with 5% milk for 1.5 h at room temperature, incubated with primary antibodies at 48C overnight and secondary antibodies at room temperature for 2 h. Bound antibodies were detected by the Pierce ECL Molecular Carcinogenesis

3

Plus Western Blotting Substrate (Thermo Fisher Scientific Inc, Santa Fe, NM) and exposed to x-ray films. Band densities were quantified by using ImagePro Plus Software (Media Cybernetics, Inc, Rockville, MD) and the densitometric results were shown. Western blot experiments were carried out in three biological replicates. Statistical Analysis The results were representative of experiments repeated at least three times and expressed as mean  SEM. Statistical comparisons were performed using Student’s t-test with two tails or ANOVA for multiple comparisons. P-values less than 0.05 were considered statistically significant ( P < 0.05). All statistical analyses were performed using Office Excel 2003 (Microsoft Corporation) or SPSS software (SPSS version 16.0). RESULTS Expression of GOLPH3 and YB1 in Human Glioma Samples To study the possible roles of GOLPH3 and YB1 in the development of human glioma, we first examined their protein levels in clinical glioma samples. As is shown in Figure 1A and B, both the protein levels of GOLPH3 and YB1 in glioma samples (38 samples) were higher than those in nontumorous tissues (19 samples). Interestingly, there was a significantly positive correlation between GOLPH3 and YB1 protein levels (Figure 1C, r ¼ 0.628). In addition, we detected GOLPH3 and YB1 expressions by immunohistochemistry (IHC) in human glioma specimens (45 samples) and human nontumorous brain tissues (20 samples) and found that both the percentage of GOLPH3 and YB1 positive cells in glioma samples were higher than that in nontumorous tissues (Figure 1D and E). These results suggest that GOLPH3 and YB1 expression are up-regulated in human glioma, which provides us the initial evidence that GOLPH3 and YB1 may play roles in the development of human glioma. Furthermore, we examined TCGA– LGG\GBM datasets and found that both GOLPH3 and YB1 mRNA levels increased significantly. Importantly, there was a direct correlation between GOLPH3 and YB1 mRNA levels (r ¼ 0.886, P < 0.01, supporting Figure 1), in line with our results found above (Figure 1C). Down-Regulation of GOLPH3 Inhibits the Migration and Invasion of Glioma Cells Because GOLPH3 markedly increased in human glioma tissues, we used the RNA interference approach to down-regulate GOLPH3 and to observe its effect on cell migration and invasion. In the preliminary experiment, we detected the basal GOLPH3 and YB1 expression in five glioma cell lines in our lab (supporting Figure 2) and found that GOLPH3 and YB1 were highly expressed in U87 and U251 cells. Thus, based on this result, we chose the U87 and U251

4

ZHANG ET AL.

Figure 1. GOLPH3 and YB1 expression in human glioma tissues. (A) Representative immunoblots of the total extracts from human glioma specimens (eight samples shown) and human nontumorous brain tissues (four samples shown). b-actin was used as a loading control. (B) Quantification of GOLPH3 and YB1 protein levels in glioma tissues (n ¼ 38) and nontumorous brain tissues (n ¼ 19) from the immunoblots. Error bars indicate  SEM.  , P < 0.01;   , P < 0.001. (C) Positive correlation between GOLPH3 and YB1 protein levels.

r ¼ 0.628, P < 0.01. (D) Representative immunohistochemistry analysis of GOLPH3 and YB1 expression in nontumorous brain tissue and glioma specimens. Nontumorous brain tissue (n ¼ 20); glioma specimens (n ¼ 45). Immunohistochemistry staining was performed in each tumor specimen at least twice with similar stained patterns. (E) The histogram showed the quantitative analysis of the percentage of GOLPH3 or YB1 positive cells. Scale bar: 50 mm. Error bars indicate  SEM.  , P < 0.001.

cell lines to perform our study. After glioblastoma U251 and U87 cell lines were transiently transfected with GOLPH3 siRNAs (si571, si635, si792), western blotting were used to examine the down-regulation efficacy. As is shown in Figures 2A–C, compared with the siRNA negative control (siNC) group, the GOLPH3 level reduced by almost 80% in the siRNA transfected cells, indicating the high down-regulation efficacy of the GOLPH3 siRNAs. To exclude the possible offtargeting effect of GOLPH3 siRNAs, the protein level of GOLPH3 paralogue GOLPH3L [30] was detected at the same time. The result showed that GOLPH3 siRNAs effectively suppress the expression of GOLPH3, but did not affect the protein expression of GOLPH3L (Figures 2A–C), suggesting the specificity of GOLPH3 siRNAs.

Thereafter, we examined the effect of GOLPH3 down-regulation on glioma cell motility by wound healing assay. We found that 24 h after scratching, siNC cells healed the wound to a greater extent than the GOLPH3 down-regulating cells (Figures 2D–G). To exclude the possibility that the effect of GOLPH3 down-regulation on cell migration is the secondary effect of GOLPH3 down-regulation on proliferation, we check whether GOLPH3 down-regulation affect glioma cell proliferation in the same culture condition. As we expected, GOLPH3 down-regulation did not show any obvious influence on glioma cell proliferation in serum-free media (supporting Figure 3A), indicating that the effect of GOLPH3 down-regulation on cell migration was not the secondary effect of proliferation. Interestingly, we

Molecular Carcinogenesis

GOLPH3-MTOR-YB1 AND GLIOMA CELL MIGRATION AND INVASION

5

Figure 2. Down-regulation of GOLPH3 inhibits the migration and invasion of glioma cells. (A–C) GOLPH3 siRNAs specifically downregulates GOLPH3 expression but not that of GOLPH3 paralogue GOLPH3L. (D, F) Down-regulation of GOLPH3 inhibits the migration and invasion in U251 and U87 cells. Cell migratory ability of different groups was detected by wound healing assay. Representative digital pictures

were taken at 0 and 24 h after scratching. Invasive ability was examined by transwell invasion assay. siNC was used as the control group. (E, G) Quantification of migratory and invasive ability of GOLPH3 downregulating cells. The migratory and invasive cell numbers of GOLPH3 down-regulating groups were normalized to that of the siNC group. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;  , P < 0.01.

observed that the cells, which migrated into the wound, were mainly the ordinary cells without being transfected by GFP-tagged GOLPH3 specific shRNA. This result indirectly demonstrated that GOLPH3 protein is up-regulated in the invading cells (supporting Figure 4). Since malignant gliomas are characterized by high invasiveness, we studied the effect of GOLPH3 on glioma cell invasion by using transwell invasion assay (matrigel precoated transwell chambers), a 3D model

of cell invasion. As is shown in Figures 2D–G, the number of invasive cells decreased markedly after GOLPH3 knockdown. These results demonstrate that down-regulation of GOLPH3 inhibits glioma cell migration and invasion.

Molecular Carcinogenesis

Over-expression of GOLPH3 Promotes the Migration and Invasion of Glioma Cells Next, we checked the effect of GOLPH3 on glioma cell migration and invasion by over-expressing

6

ZHANG ET AL.

Figure 3. Over-expression of GOLPH3 promotes the migration and invasion of glioma cells. (A) Representative immunoblots of the total extracts from MYC-GOLPH3 transiently transfected glioma or control cells. (B) Cell migratory ability of different groups was detected by wound healing assay. Representative digital pictures were taken at 0 and 24 h after scratching. Invasive ability was examined by

Molecular Carcinogenesis

transwell invasion assay. Empty vector was used as the control. (C, D) Quantification of migratory and invasive ability of GOLPH3 overexpression cells. The migratory and invasive cell numbers of GOLPH3 over-expression group were normalized to that of the control group. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;  , P < 0.01.

GOLPH3-MTOR-YB1 AND GLIOMA CELL MIGRATION AND INVASION

7

What is the mechanism of GOLPH3 on glioma cell migration and invasion? mTOR is a serine/threonine kinase that regulates cell growth, proliferation, survival, metabolism, and angiogenesis by integrating signals received from the PI3K/Akt signaling path-

way [31]. Aside from PI3K and AKT, other key signaling molecules upstream of mTOR include GOLPH3 [11]. After being activated, mTOR activates its downstream targets ribosomal protein S6 kinase 1 (S6K1) and the eukaryotic protein synthesis initiation factor 4E-binding protein 1 (4EBP1, also called EIF4EBP1) [31] to exhibit its functions. Further study found that mTOR also regulates the invasion and metastasis of prostate cancer through its downstream target gene YB1 [32]. Our previous study showed that GOLPH3 can regulate migration and invasion of glioma cells [15]. Thus, we guess whether the effect of GOLPH3 on glioma migration and invasion is through regulating mTOR and YB1. Firstly, we tested the possibility that GOLPH3 regulates the mTOR activity. As is shown in Figures 4A and B, compared with siNC group, the protein levels of p-p70S6K and p-4EBP1, two classic downstream molecules reflecting the activity of mTOR, decreased significantly after GOLPH3 downregulation. The levels of p-p70S6K and p-4EBP1 also

Figure 4. GOLPH3 regulates mTOR activity and mTOR inhibitor INK128 inhibits glioma cell migration and invasion. (A, B) Downregulation of GOLPH3 decreased the levels of p-p70S6K, p-4EBP1 and p-YB1, as well as the level of YB1. (C, D) Compared with the control group, mTOR inhibitor INK128 (200nM) decreased the levels of p-p70S6K and p-4EBP1, while rapamycin (Rapa 50nM) treatment did not decrease the level of p-4EBP1. In addition, suppression of mTOR activity by INK128 decreased the protein level of YB1. (E-G)

Suppression of mTOR activity by inhibitor INK128 inhibited glioma cell migration and invasion. Cell migratory ability of different groups was detected by wound healing assay. Representative digital pictures were taken at 0 and 24 h after scratching. Invasive ability was examined by transwell invasion assay. The migratory and invasive cell numbers of rapamycin or INK128 treatment group were normalized to that of the control group. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;  , P < 0.01.

GOLPH3. As is shown in Figure 3A, the exo-GOLPH3 expressed very well after MYC-GOLPH3 transfection. The wound healing assay revealed that the wound healed obviously and had the fusion tendency after over-expression of GOLPH3 (Figures 3B–D). Similar to the effect of GOLPH3 down-regulation, over-expression of GOLPH3 did not affect glioma cell proliferation in serum-free media (supporting Figure 3B). The matrigel invasion assay revealed that the invasiveness of glioma cells enhanced significantly too (Figures 3B–D). These results demonstrate that overexpression of GOLPH3 promote glioma cell migration and invasion. GOLPH3 Regulates mTOR Activity and mTOR Inhibition Reduces the Migration and Invasion of Glioma Cells

Molecular Carcinogenesis

8

ZHANG ET AL.

Figure 4. (Continued )

decreased after INK128 treatment in U251 cells. INK128 is an mTOR ATP site inhibitor. On the contrary, rapamycin, an allosteric mTOR inhibitor, only decreased p-p70S6K level in these cells (Figures 4C and D). Consistently, suppression of mTOR activity by INK128 inhibited glioma U251 and U87 cell migration and invasion, while rapamycin showed no effects (Figures 4E–G). Furthermore, both INK128 and rapamycin did not affect glioma cell proliferation in serum-free media (supporting Figure 3C), indicating that the effect on cell migration was not the secondary effect of proliferation. These results indicate that GOLPH3 can regulate the activity of mTOR and suppression of mTOR inhibits glioma cell migration and invasion. GOLPH3 Regulates YB1 and YB1 Down-Regulation Reduces Migration and Invasion of Glioma Cells Next, we tested the possibility of GOLPH3 regulating YB1, one of mTOR target genes taking responsibility for cell migration and invasion [32]. As is shown in Figures 5A and B, the protein level of YB1 decreased after GOLPH3 si792 transfection, in line with the result obtained in Figures 4A and B. Furthermore, GOLPH3 down-regulation also decreased p-YB1 level, suggesting that GOLPH3 also regulate YB1 activation Molecular Carcinogenesis

(Figure 4A). On the contrary, the protein level of YB1 increased after GOLPH3 over-expression (Figures 5C and D). These results indicate that GOLPH3 regulate YB1 protein level and activity. Next, we tested whether YB1 has the similar effect on the migration and invasion of glioma cells to that of GOLPH3. Firstly, we down-regulated the YB1 by RNA interference approach and examined the downregulation efficacy by western blotting. After glioma U251 cells were transiently transfected with YB1 siRNA, the protein level of YB1 decreased dramatically, suggesting high down-regulation efficacy of YB1 siRNA (Figures 5E and F). However, when YB1 was knocked down, no change of GOLPH3 expression was observed (Figures 5E and F), suggesting that GOLPH3 is located at the upstream of YB1 and regulate the expression of YB1 in glioma U251 cells. Thereafter, the wound healing assay and matrigel invasion assay were used to test the effect of YB1 on glioma cell migration and invasion. As the Figures 5G–I showed, both the number of migratory and invasive glioma cells significantly reduced after YB1 siRNA transfection compared with the siNC group. Furthermore, down-regulation of YB1 did not affect glioma cell proliferation in serum-free media (supporting Figure 3D), indicating that the effect on cell migration was not the secondary effect of proliferation.

GOLPH3-MTOR-YB1 AND GLIOMA CELL MIGRATION AND INVASION

9

Figure 5. GOLPH3 regulates YB1 and YB1 down-regulation reduces migration and invasion of glioma cells. (A, B) The level of YB1 decreased after GOLPH3 down-regulation. YB1 levels were normalized to that of b-actin. (C, D) The level of YB1 increased after GOLPH3 overexpression. (E, F) Compared with the siNC group, down-regulation of YB1 did not change the level of GOLPH3. (G) Effect of YB1 downregulation on U251 and U87 cell migration was examined by wound

healing assay. Representative digital pictures were taken at 0 and 24 h after scratching. Effect of YB1 down-regulation on cell invasive ability was examined by transwell invasion assay. Scale bar, 100 mm. (H, I) Quantification of migratory and invasive ability of YB1 down-regulation cells. The migratory and invasive cell numbers of YB1 down-regulation group were normalized to that of the control group. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;   , P < 0.01.

GOLPH3 Promotes Glioblastoma Cell Migration and Invasion via mTOR-YB1 Pathway

ing [11]. Later studies further demonstrate that GOLPH3 plays important roles in the progression of many tumors [13,14]. In our previous study, we found that GOLPH3 down-regulation inhibited glioma cell migration and invasion [15]. However, how GOLPH3 regulates cell migration and invasion is largely unknown. In this study, we provided the evidence that GOLPH3-mTOR-YB1 signal pathway was involved in glioma cell migration and invasion in vitro. Several lines of results support this conclusion. First, GOLPH3 and YB1, both up-regulated in human glioma tissues, show direct correlation with each other. Second, down-regulation of GOLPH3 inhibited glioma cell migration and invasion, while overexpression of GOLPH3 enhanced them. Third, YB1 level and mTOR activity decreased after GOLPH3 down-regulation but increased after GOLPH3 overexpression. Forth, YB1 down-regulation or mTOR inhibitor INK128 treatment inhibited cell migration and invasion, similar to the effect of GOLPH3 downregulation. Finally, over-expression of GOLPH3 induced glioma cell migration and invasion was blocked by INK128 treatment and YB1 down-regulation. The mTOR kinase is a master regulator of protein synthesis that couples nutrient sensing with cell growth and cancer. The major regulators of protein

The above data indicate that both GOLPH3 and mTOR affect the migratory and invasive behavior of glioma cells and GOLPH3 regulate mTOR activity and YB1 expression. Therefore, we examined whether mTOR or YB1 could mediate the function of GOLPH3 in glioma cell migration and invasion. Figures 6A– C, 7A, and B showed that over-expression of GOLPH3 increases the protein level of YB1 and/or p-4EBP1, consistant with the results of Figure 4A and B. Interestingly, the above effects were abolished by mTOR inhibitor INK128 treatment or YB1 downregulation, suggesting that GOLPH3 regulates YB1 via mTOR pathway. In addition, GOLPH3 over-expression induced glioma cell migration and invasion increase were abolished by INK128 treatment and YB1 down-regulation (Figures 6D–F and 7C–E), indicating that GOLPH3 regulates YB1 promoting glioblastoma cell migration and invasion via the mTOR pathway (Figure 7F). DISCUSSION Recently, GOLPH3 was identified as a novel oncogene promoting cell transformation and tumor growth by constitutively activating mTOR signalMolecular Carcinogenesis

10

ZHANG ET AL.

Figure 6. Over-expression of GOLPH3 induced glioma cell migration and invasion is blocked by mTOR specific inhibitor INK128. (A–C) MYCGOLPH3 was transfected successfully into glioma U251 cells and inhibitor INK128 reduced the protein level of p-4EBP1. After GOLPH3 over-expression, the cells were treated with or without INK128 (200 nM) and YB1 protein level was determined by western blotting.

(D–F) Over-expression of GOLPH3 induced glioma cell migration and invasion was blocked by mTOR specific inhibitor INK128. After GOLPH3 over-expression, the cells were treated with or without INK128 and the migratory and invasive ability were detected by wound healing assay or transwell invasion assay respectively. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;  , P < 0.01.

synthesis downstream of mTOR are 4EBP1 and p70S6K [33,34]. 4EBP1 negatively regulates eIF4E, a key rate-limiting initiation factor for cap-dependent translation. Phosphorylation of 4EBP1 by mTOR leads to its dissociation from eIF4E, allowing translation initiation complex formation at the 50 end of mRNAs [35]. The mTOR-dependent phosphorylation of p70S6K also promotes translation initiation as well as elongation [36]. In this study, we found that the phosphorylation of p70S6K and 4EBP1 decreased after INK128 treatment, while the protein level of p-4EBP1 showed no difference after rapamycin treatment (Figures 4C and D). Furthermore, suppression of mTOR activity by INK128, but not by rapamycin, inhibited glioma cells migration and invasion, consistent with the previous studies [32]. These results indicate that suppression of mTOR reduce glioma cell migration and invasion. Andrew C. Hsieh also found that INK128, but not rapamycin, decreases the invasive potential of PC3 prostate cancer cells.

However, Steffen Wedel discovered that rapamycin reduced prostate cancer cell adhesion, migration and invasion, accompanied by elevated Akt activation and p70S6 kinase de-activation [37]. Perhaps the roles of rapamycin in tumor cell migration and invasion depend on drug concentration or tissue specificity. In addition, we found that down-regulation of GOLPH3 decreased the phosphorylation of p70S6K and 4EBP1 (Figures 4A and B), in line with the previous study that mTOR was constitutively activated by GOLPH3 [11]. Furthermore, the effect of GOLPH3 over-expression on cell migration and invasion were abolished by mTOR ATP site inhibitor INK128 treatment (Figures 6D–F). Our study shows that GOLPH3 regulate mTOR activity not only to promote cell growth but also to enhance cell motility. The mTOR downstream translationally regulated nodes of gene expression that may direct cancer invasion and metastasis are poorly characterized. Andrew C. Hsieh used ribosome profiling to generate

Molecular Carcinogenesis

GOLPH3-MTOR-YB1 AND GLIOMA CELL MIGRATION AND INVASION

11

Figure 7. Over-expression of GOLPH3 induced glioma cell migration and invasion is blocked byYB1 downregulation. (A, B) Glioma cells were co-transfected with MYC-GOLPH3 and YB1 siRNA. YB1 levels were evaluated by western blotting. (C–E) Over-expression of GOLPH3 induced glioma cell migration and invasion was blocked byYB1 down-regulation. Scale bar, 100 mm. Error bars indicate  SEM.  , P < 0.05;  , P < 0.01. (F) Proposed model of GOLPH3-mTOR-YB1 signaling pathway in glioblastoma cells.

a comprehensive map of translationally controlled mTOR targets in cancer that surprisingly stratify into specific cellular processes including proliferation, metabolism, protein synthesis and invasion [32]. Among mTOR targets, YB1 takes the responsibility for cell migration and invasion. In this study, we found that YB1 down-regulation inhibits glioma cell migration and invasion (Figures 5G–I), which is consistent with the recent reports on translational activation of snail1 and other developmentally regulated transcription factors by YB1 promoting an epithelial-mesenchymal transition [26]. GOLPH3 down-regulation decreased not only the protein level of YB1 but also the phosphorylation of YB1 (Figures 4A and B, 5A and B). On the contrary, overexpression of GOLPH3 increased the protein level of YB1 (Figure 5C and D). Furthermore, the effect of GOLPH3 over-expression on cell migration and invasion were abolished by YB1 down-regulation (Figure 7C–E). However, no change of GOLPH3 expression was observed after YB1 down-regulation (Figure 5E and F), suggesting that GOLPH3 is located at the upstream of YB1. Our results indicated that GOLPH3 regulate the activity of mTOR and the Molecular Carcinogenesis

expression and activity of YB1, which performs its effect on glioma cell migration and invasion. Our study links GOLPH3 to the poorly understood mechanisms underlying migration and invasion of glioma cells via the mTOR pathway. Several questions need to be addressed in future study. Firstly, the result of western blotting showed that it is mTOR ATP site inhibitor INK128, but not rapamycin, that decreased the level of p-4EBP1 and YB1 at the same time. In other words, altered phosphorylation of the mTOR substrate 4EBP1 affected the protein level of YB1, which is not achieved by altering phosphorylation of the other mTOR substrate S6K. Thus, the 4EBP1-eIF4E axis is probably druggable with potent mTOR ATP site inhibitor INK128, which is different from rapamycin and targets mTOR-dependent 4EBP1 phosphorylation. However, the role of p-4EBP1, the bridge of GOLPH3 and YB1, requires further study. Secondly, as we know, YB1 are ubiquitously expressed and involved in fundamental processes such as DNA repair, mRNA transcription, splicing, translation and stabilization [16,17], but the role and mechanism of YB1 in glioma cell migration and invasion are still unknown. Karen reported that the expression of

12

ZHANG ET AL.

CD44 correlates with the presence of p-YB1S102 in SUM 149 cells in vitro and in fixed tumour xenograft [24]. Furthermore, activated mutant YB1S102D enhances self-renewal, primary and secondary mammosphere growth, and soft agar colony growth in human breast cancer cell lines [24]. Therefore, further studies of the mechanism of YB1 promoting glioblastoma cell migration and invasion are required.

REFERENCES 1. Huse JT, Holland EC. Targeting brain cancer: Advances in the molecular pathology of malignant glioma and medulloblastoma. Nat Rev Cancer 2010;10:319–331. 2. Clarke J, Butowski N, Chang S. Recent advances in therapy for glioblastoma. Arch Neurol 2010;67:279–283. 3. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med 2008;359:492–507. 4. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352:987–996. 5. Manji HK, Quiroz JA, Sporn J, et al. Enhancing neuronal plasticity and cellular resilience to develop novel, improved therapeutics for difficult-to-treat depression. Biol Psychiatry 2003;53:707–742. 6. Snyder CM, Mardones GA, Ladinsky MS, Howell KE. GMx33 associates with the trans-Golgi matrix in a dynamic manner and sorts within tubules exiting the Golgi. Mol Biol Cell 2006;17:511–524. 7. Dippold HC, Ng MM, Farber-Katz SE, et al. GOLPH3 bridges phosphatidylinositol-4-phosphate and actomyosin to stretch and shape the Golgi to promote budding. Cell 2009;139:337– 351. 8. Wood CS, Schmitz KR, Bessman NJ, Setty TG, Ferguson KM, Burd CG. PtdIns4P recognition by Vps74/GOLPH3 links PtdIns 4-kinase signaling to retrograde Golgi trafficking. J Cell Biol 2009;187:967–975. 9. Schmitz KR, Liu J, Li S, et al. Golgi localization of glycosyltransferases requires a Vps74p oligomer. Dev Cell 2008;14:523–534. 10. Tu L, Tai WC, Chen L, Banfield DK. Signal-mediated dynamic retention of glycosyltransferases in the Golgi. Science 2008; 321:404–407. 11. Scott KL, Kabbarah O, Liang MC, et al. GOLPH3 modulates mTOR signalling and rapamycin sensitivity in cancer. Nature 2009;459:1085–1090. 12. Sabatini DM. mTOR and cancer: Insights into a complex relationship. Nat Rev Cancer 2006;6:729–734. 13. Kunigou O, Nagao H, Kawabata N, et al. Role of GOLPH3 and GOLPH3L in the proliferation of human rhabdomyosarcoma. Oncol Rep 2011;26:1337–1342. 14. Li XY, Liu W, Chen SF, Zhang LQ, Li XG, Wang LX. Expression of the Golgi phosphoprotein-3 gene in human gliomas: A pilot study. J Neurooncol 2011;105:159–163. 15. Zhou X, Zhan W, Bian W, et al. GOLPH3 regulates the migration and invasion of glioma cells though RhoA. Biochem Biophys Res Commun 2013;433:338–344. 16. Evdokimova VM, Ovchinnikov LP. Translational regulation by Y-box transcription factor: Involvement of the major mRNAassociated protein, p50. Int J Biochem Cell Biol 1999;31: 139–149. 17. Kohno K, Izumi H, Uchiumi T, Ashizuka M, Kuwano M. The pleiotropic functions of the Y-box-binding protein, YB-1. Bioessays 2003;25:691–698. 18. Lyabin DN, Eliseeva IA, Ovchinnikov LP. YB-1 Synthesis is regulated by mTOR signaling pathway. PLoS ONE 2012;7: 1–11.

Molecular Carcinogenesis

19. Stein U, Jürchott K, Walther W, Bergmann S, Schlag PM, Royer HD. Hyperthermia-induced nuclear translocation of transcription factor YB-1 leads to enhanced expression of multidrug resistance-related ABC transporters. J Biol Chem 2001;276: 28562–28569. 20. Ohga T, Koike K, Ono M, et al. Role of the human Y boxbinding protein YB-1 in cellular sensitivity to the DNAdamaging agents cisplatin, mitomycin C, and ultraviolet light. Cancer Res 1996;56:4224–4228. 21. Bargou RC, Jürchott K, Wagener C, et al. Nuclear localization and increased levels of transcription factor YB-1 in primary human breast cancers are associated with intrinsic MDR1 gene expression. Nat Med 1997;3:447–450. 22. Kuwano M, Oda Y, Izumi H, et al. The role of nuclear Y-box binding protein 1 as a global marker in drug resistance. Mol Cancer Ther 2004;3:1485–1492. 23. Knösel T, Emde A, Schlüns K, et al. Immunoprofiles of 11 biomarkers using tissue microarrays identify prognostic subgroups in colorectal cancer. Neoplasia 2005;7:741–747. 24. To K, Fotovati A, Reipas KM, et al. Y-box binding protein-1 induces the expression of CD44 and CD49f leading to enhanced self-renewal, mammosphere growth, and drug resistance. Cancer Res 2010;70:2840–2851. 25. Bader AG, Vogt PK. Inhibition of protein synthesis by Y boxbinding protein 1 blocks oncogenic cell transformation. Mol Cell Biol 2005;25:2095–2106. 26. Evdokimova V, Tognon C, Ng T, et al. Translational activation of snail1 and other developmentally regulated transcription factors by YB-1 promotes an epithelial-mesenchymal transition. Cancer Cell 2009;15:402–415. 27. Thiery JP, Sleeman JP. Complex networks orchestrate epithelialmesenchymal transitions. Nat Rev Mol Cell Biol 2006;7:131–142. 28. Janz M, Harbeck N, Dettmar P, et al. Y-box factor YB-1 predicts drug resistance and patient outcome in breast cancer independent of clinically relevant tumor biologic factors HER2, uPA and PAI-1. Int J Cancer 2002;97:278–282. 29. Bergmann S, Royer-Pokora B, Fietze E, et al. YB-1 provokes breast cancer through the induction of chromosomal instability that emerges from mitotic failure and centrosome amplification. Cancer Res 2005;65:4078–4087. 30. Ng MM, Dippold HC, Buschman MD, Noakes CJ, Field SJ. GOLPH3L antagonizes GOLPH3 to determine Golgi morphology. Mol Biol Cell 2013;24:796–808. 31. Fasolo A, Sessa C. Targeting mTOR pathways in human malignancies. Curr Pharm Des 2012;18:2766–2777. 32. Hsieh AC, Liu Y, Edlind MP, et al. The translational landscape of mTOR signalling steers cancer initiation and metastasis. Nature 2012;485:55–61. 33. Brown EJ, Beal PA, Keith CT, Chen J, Shin TB, Schreiber SL. Control of p70 s6 kinase by kinase activity of FRAP in vivo. Nature 1995;377:441–446. 34. Gingras AC, Kennedy SG, O'Leary MA, Sonenberg N, Hay N. 4E–BP1, a repressor of mRNA translation, is phosphorylated and inactivated by the Akt(PKB) signaling pathway. Genes 1998;12:502–513. 35. Gingras AC, Raught B, Sonenberg N. Regulation of translation initiation by FRAP/mTOR. Genes Dev 2001;15:807–826. 36. Ruvinsky I, Meyuhas O. Ribosomal protein S6 phosphorylation: from protein synthesis to cell size. Trends Biochem Sci 2006; 31:342–348. 37. Wedel S, Hudak L, Seibel JM, et al. Impact of combined HDAC and mTOR inhibition on adhesion, migration and invasion of prostate cancer cells. Clin Exp Metastasis 2011;28: 479–491.

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site.