Sorcin induces gastric cancer cell migration and invasion contributing to STAT3 activation

Gastric cancer (GC) is a globally occurring malignancy that is characterized by a high mortality rate due to a high tendency to metastasize and poor prognoses. Sorcin, as known as SRI, a soluble resistance-related calcium-binding protein, plays a significant role in multidrug resistance. Sorcin is related to the migration and invasion of cancer cells. However, the mechanism remains unclear. Here, we used immunohistochemistry to confirm that the expression of sorcin in cancer tissues is higher than that in the adjacent normal tissues. The wound healing and transwell results indicate that sorcin can induce migration and invasion of GC cells. To explore the role of sorcin in GC metastasis, isobaric tags for relative and absolutely quantitation (iTRAQ) were used to examine cells with and without sorcin knockdown to identify the differentially expressed proteins (DEPs). The results were evaluated via RT-PCR and western blot to confirm the ITRAQ data. Inhibition of sorcin expression can down- regulate the expression of CTSZ, MMP2, MMP9 and p-STAT3 followed by suppression of tumor growth and metastasis. Together, we concluded that sorcin has a oncogenic activity via inducing tumor growth and metastasis, leading to development of therapeutic treatments for GC.


INTRODUCTION
Gastric cancer is the second leading cause of cancerassociated death worldwide [1,2]. Approximately 40% of all GC cases occur in China and these are often diagnosed in the advanced stages [3,4]. The 5 year survival rate was 28% in 2014 [5]. Many factors result in the occurrence of GC, including genetic and epigenetic alterations. In spite of this, tumor suppression genes and growth factors have been discovered which affect the progression of GC [6][7][8], although the molecular mechanism of tumor metastasis is still poorly documented [9].
Sorcin, a soluble resistance-related calcium-binding protein, belongs to the small penta-EF-hand protein family [10,11]. Sorcin expressed in many human tissues and at high levels in bones, heart, brain, kidneys, breasts and skin. Sorcin has been found to over expressed in many cancers such as leukemia and gastric, ovarian and breast cancers [10,12,13]. Although the function of sorcin in tumors is still unclear, a number of studies have shown www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 61), pp: 104258-104271 Research Paper that sorcin may associate with multidrug resistance (MDR) [12,[14][15][16]. Using a gene co-expressed with P-glycoproteins in multidrug-resistant cells, sorcin was shown to be resistance-related [17]. On the other hand, sorcin is related to epithelial-to-mesenchymal (EMT) transition, which contributes to cancer metastases, as in breast cancer and colorectal cancer [14,18].
Since sorcin plays a significant role in cancer metastasis and the mechanism for this action remains elusive, the present study utilized isobaric tags for lative and absolute quantitation (iTRAQ) to identify differentially expressed proteins (DEP) in AGS and MKN-28 gastric cancer cell lines in which sorcin had been silenced via Small interfering RNA (siRNA). After verification by PT-PCR and western blot, the DEPs were further investigated in order to identify a pathway to explain the sorcin-modulated metastatic mechanism. Our data demonstrated that sorcin can induce tumor growth and metastasis via STAT3 signaling, leading to development of therapeutic treatments for GC.

Differential expression of sorcin in tissues
IHC was utilized to verify whether the expression of sorcin differed between tumor tissues and adjacent normal tissues. The results showed that sorcin expression was higher in tumor tissues than that in normal adjacent tissues ( Figure 1A). The IHC value of sorcin in tumor tissues was significantly higher than that in normal tissues ( Figure  1B). Moreover, Pearson correlation analysis showed that IHC scoring was correlated with the Grade and Stage of tumor ( Table 1).

Effect of sorcin knockdown on migration and invasion in AGS and MKN28 cell lines
In order to identify the association of sorcin with GC migration and invasion, the control siRNA and SRI siRNA (SRI siRNA1 and SRI siRNA3) were transfected into the AGS and MKN28 cell lines. Cells transfected with the SRI siRNA showed significant suppression of sorcin expression (Figure 2A). Pursuant with these results, the invasion and migration capabilities of AGS and MKN28 cells were significantly decreased after transfection with SRI siRNA (Figure 2B and 2C).

Analysis of iTRAQ data of differentially expressed proteins
The iTRAQ technology was used to explore the mechanism by which sorcin affects cell migration and invasion. The iTRAQ assays were performed on proteins isolated from AGS and MKN28 cells sorcin knockdown vs control cells without sorcin silencing. Specimens were labeled in duplicate to improve the confidence and enhance the range of protein identification. Table 2 shows the flow chart of the iTRAQ proteomics approach. The ratios of 114:113, 116:115, 118:117 and 121:119 identified the differential protein expression. An additional 1.3-fold change cutoff for all iTRAQ ratios (ratio <0.77 or >1.3) was used for classifying proteins as being up-or downregulated.
A total of 65 proteins were identified in AGS cells and 57 proteins in MKN28 cells. According to the conventions, 22 proteins were up-regulated and 43 proteins were down-regulated in the AGS line. In the MKN-28 line, 25 proteins were up-regulated and 32 were down-regulated (Table 3). To identify the cellular and molecular characteristics of these proteins, the differentially expressed proteins were classified using the PANTHER (www.pantherdb.org/) Classification System ( Figure 3A and 3B). Using STRING analysis, sorcin was identified as the most important node in the DEP network because it had the greatest connectivity ( Figure 3C).

Validation of iTRAQ identified candidate proteins
PT-PCR and western blot were used to confirm the changes in the selected proteins. mRNA expression of UBE2T, CTSZ, HSPBP1, S100A11, CSTB, MMP1 and RAC1 were in line with the iTRAQ resultsin AGS cells. In the MKN-28 cell line, mRNA expression of UBE2T, CTSZ, HSPBP1, KRT1, TIMP2, VEGFA and IBP4 were similar to the iTRAQ results. After silencing of sorcin, the mRNA expression of UBE2T was significantly upregulated compared with 18s, while the mRNA expression of the other identified proteins was down-regulated ( Figure 4). Western blot analysis was utilized to quantify the expression levels of the selected proteins ( Figure 5A and 5B). The results of CTSZ and HSPBP1 are consistent to the iTRAQ data.
Knockdown of sorcin suppresses p-STAT3, MMP9 and MMP2 protein expression. Many studies have reported that the overexpression of p-STAT3, MMP2 and MMP9 promotes metastasis. We detected the expression of these proteins in both silenced and untreated cells. The results demonstrated that the down regulation of sorcin inhibits p-STAT3, MMP2, MMP9 expression at the protein level in AGS and MKN28 cells (p<0.05) ( Figure  5A and 5B).

Sorcin plays a role in proliferation and cell cycle
AGS and MKN28 cells were transfected with sorcin siRNA sequences. According to the western blot analysis, efficient silencing of sorcin expression was demonstrated by the sorcin-specific siRNA sequences. We also used the MTT assay to examine the proliferation of sorcin-silenced vs control AGS cells and MKN-28 cells. The proliferation of AGS and MKN28 cells was depressed compared to the control cells. We examined the change of the cell cycle and the results showed a difference in the cycles of sorcinsilenced cells compared with untreated cells ( Figure 6A and 6B).

DISCUSSION
Gastric cancer has high mortality and mobility rate and is the third leading cause of cancer deaths [19,20]. Surgery and chemotherapy are the prime treatment for curing the malignancy [8,20]. According to official  3 Stage 0.869 ** 0.92 ** 1 * P < 0.05 ** P < 0.01 *** P < 0.001.   statistics, about 50% of GC cases suffer metastases at diagnosis [21]. The 5-year survival rate is greatly decreased in patients with metastases compared with patients diagnosed early in the progression [20]. Therefore, it is urgent to explore the molecular mechanisms underlying the progression of GC, which could identify biomarkers or treatments for GC. Sorcin is over expressed in many malignant tumors. In the present study, the iTRAQ method was utilized to compare protein expression in GC tumor tissue and www.impactjournals.com/oncotarget compare them with expression from adjacent tissues. As previously reported, sorcin plays a vital role in invasion and migration of tumors, also has an influence on tumor proliferation. Some of DEPs identified by the iTRAQ results were attributed to cellular processes. CTSZ, HspBp1, S100A11, CTSB, UBET2E, KRT1, MMP1, Rac1 and VEGFA were selected as candidate proteins, and PT-PCR and western blot analyses were used to validate the iTRAQ results. To identify the major changes after suppressing sorcin, we classified these proteins into different categories according to their biological process and molecular functions.
CTSZ is considered a lysosomal proteolytic enzyme and is a member of the peptide C1 family associated with many biological processes, including immune response, cell adhesion and proliferation [22]. CTSZ is over expressed in many cancers [23,24]. CTSZ belongs to the cathepsin family of lysosomal hydrolases that contribute to the turnover of intra cellular proteins and the degradation of extracellular matrix [25]. Interestingly, several of the tumor-promoting functions of CTSZ were not dependent on its described catalytic activity [26]. Several proteases have been shown to mediate adhesion and migration of cells through interaction with integrins [27]. CTSZ contains an Arg-Gly-Asp (RGD) motif allows it to interact with integrins [26,28]. RGD is a protein sequence vital to the binding of proteins to cell surface [29], which mediates cell adhesive properties, and this region interacts with the integrin αvb3 [30]. This interaction influences cellular migration as well as cell adhesion [30,31]. Many studies have shown that CTSZ is related with the EMT, which could promote invasion and migration in tumor [32]. The extracellular matrix (EMT) plays an important role in modulating cell behavior, including cell survival and metastasis [33].
MMPs, a group of zinc-dependent endopeptidases, are involved in degradation of the ECM. MMP2 and MMP9 contribute to tumor metastasis [34][35][36]. Studies have reported signal transducer and activator of transcription 3 (STAT3) is active in many cancers [37][38][39][40][41]. In the present study, we found that silencing sorcin was related to metastatic activity in tumor cells. Therefore, we suspect that sorcin may be associated with the STAT3 pathway. We found that p-STAT3, MMP2, MMP9 were suppressed after silencing sorcin in GC cell lines. Taken together, these findings suggest that sorcin contribute to STAT3 activation.
In conclusion, our findings indicated that silencing sorcin affects the development of GC cells, by reducing the capacity for invasion, migration and proliferation via

Cell lines
The two human gastric cancer lines AGS and MKN28 were obtained from ATCC, and grown in high glucose DMEM with 10% fetal bovine serum (FBS Corning), augmented with 1% penicillin and streptomycin. Cells were maintained in an incubator in an atmosphere of 5% carbon dioxide at 37°C.

Sample collection and ITRAQ labeling
Cells were grown to 80% confluence, washed two times with PBS and incubated in DMEM for 24h. The supernatant containing secreted proteins was collected and filtered using a 0.22 mm filter (Millipore, Bedford, MA, USA) and concentrated by centrifugation with an Amican centrifugal filter (Billerica, MA, USA). The protein concentration was measured using a 2D Quant kit. For each sample, 100 ug of protein was precipitated overnight, re-dissolved in lysis buffer, and cysteine-blocked in accordance with the manufacturer's instructions. Th protein samples were then labeled with the iTRAQ tags. The secretory protein samples from AGS without sorcin knockdown were labeled with tag 113 and tag 115. The secretory proteins from AGS cells with sorcin-silencing were labeled with tag 114 and tag 116. The secretory proteins from MKN-28 cells without sorcin knockdown were labeled with tags 117 and 119 and the secretory proteins from sorcin-silenced MKN-28 cells were labeled with tags 118 and 121.

Fractionation of pepitides
The pooled iTRAQ labeled samples were dissolved in 300uL 1% Pharmalyte (Amersham Biosciences) and 8M urea and evenly coated on pre-hydrated IPG strips (pH 3-8). The peptides were isolectrically focused successively for 1h at 500v, 1h at 1000v, 1h at 3000v, and 8.5h at 8000v to give a total of 68kv in an IPGphor system (Amersham Biosciences). The peptides were extracted using an acetonitrile (ACN) and formic acid solution. Fractions were lyophilized in a vacuum concentrator and purified on an SPE C18 column. The purified fractions were re-lyophilized and stored in -20°C for use in mass spectrometric analysis.

Mass spectrometry
The purified, labeled peptides were reconstituted in Buffer A (2% ACN; 0.1% formic acid) and injected into the nano LC ESI MS/MS system. Mass spectrometry was performed using a Qstar Elite mass spectrometer coupled to a Dionex Ultimate 3000 liquid chromatography system. A gradient series for each analysis was loaded on C-18 pepMap column at a flow rate of 0.3ul/min. The mass spectrometer was set to perform data acquisition in the positive ion mode, with a selected mass range of 300-1800 m/z. The two most abundantly charged ions above 20 counts were selected for MS/MS. Dynamic exclusion criteria was set to 30s with a ± 50 Da mass tolerance. Protein Pilot software (version 2.0 applied biosystems, MDS Sciex) was used for the identification and qualification of proteins, with the search being carried out against the International Protein Index (IPI) human database (version 3.77).
Immunohistochemistry and tissue Microarray IHC assessment of sorcin was performed using a commercial tissue microarray (Biomax, Rockville, MD, USA) containing 40 GC samples and 40 matched GC adjacent normal tissue. Immunohistochemistry was performed on the tissue microarrays as previously described [42]. The tissue samples were dewaxed with xylene and rehydrated using an alcohol gradient. The samples were washed three times in double distilled H 2 O (5 min per wash) and subjected to heat-induced antigen retrieval in a 0.1 M citrate solution for 5 minutes. Endogenous peroxidase activity in the samples was quenched by incubation in 3% H 2 O 2 . The sections were blocked with BSA for 0.5h and incubated with anti-sorcin antibody overnight at 4°C. Detection was performed on an Envision/horseradish peroxidase system (DakoCytomation, Glostrup, Denmark) and all slides were counterstained with Gill hematoxylin for examination. Protein expression was evaluated by recording staining intensity (scale 0-3) and the percentage of positive cells (0-100%). Multiplying these scores yielded a value of 0-300. All data analyses were performed with SPSS v.16.0 software (SPSS, Chicago, IL, USA) using the student the t-test with 95% confidence levels.

Sorcin siRNA transfection, wound healing, cell migration and invasion assays
AGS and MKN28 cells were transfected with 50nm of sorcin specific Stealth Selected RNAi siRNA (HSS) or a negative control siRNA using Lipofactamin 2000, according to the manufacturers protocol (Life Technologies, Carlsbad, CA, USA). The cells were cultured in serum-free, high-glucose DMEM with free serum for 2 days in 6-well plates. When the cells became confluent, a scratch was made in the monolayer with a sterile P200 pipette tip. The cells were washed twice with PBS to remove debris and the wound channel was photographed under a phase-contrast microscope at 0h and 24h for comparison. Adobe Photoshop 7.0 was used to measure the relative width of the wound, and cell migration was determined by the ability of the cells to close the scratch area. Migration and invasion assays was performed using Cell Invasion Assay kits (Cell Biolabs, Inc., Beijing China). AGS and MKN28 cells were transfected with negative siRNA or sorcintargeting siRNA and cultured in free serum media. Approximately 3 * 10 5 /300ul cells were seeded into the upper chambers of the Transwell chambers, and 500ul media was loaded into the lower chamber. The migration assay cells were cultured for 12h, while the invasion assay cells were cultured for 24h. The non-invasive cells were gently removed with cotton and the invading cells were stained, fixed, extracted and quantified with cyQuant GR fluorescent dye at 560 nm. The knockdown of sorcin protein expression was confirmed by western blot.

Western blot analysis
Cells were lysed with lysis buffer and the concentration of the proteins was determined by a 2D Quantification kit (GE Healthcare). The proteins were separated by 12% SDS-PAGE and transfected to PVDF membranes. The membranes were blocked in BSA for 1h at room temperature, and incubated with the primary antibody (1:500 dilution) overnight at 4°C. The membranes were washed three times and incubated with HRP-conjugated secondary antibody (1:10000). The membranes were washed thrice with TBST and visualized with the ChemiDoc MP imaging system (BioRad Laboratories, Hercules, CA, USA).

Cell proliferation assay
AGS and MKN28 cells were seeded in 96-well plates at a density of 1.5x10 3 cells/well. Cells were cultured in DMEM with 10% FBS and transfected with SRI siRNA or control siRNA for 0, 24, 48, 72 and 96 h at 37˚C. The MTT assay was performed as follows: cells were incubated with 20 μl MTT (Sigma-Aldrich, St Louis, USA) at 37˚C for 4 h. The MTT substrate was then dissolved in 200 μl of DMSO (Sigma-Aldrich, St Louis, USA) for 5 min. The absorbance was measured at 570 nm.

Statistical analysis
All experiments were performed in triplicate and the data are expresses as the mean ± standard deviation. The Student's t-test was used to analysis the variables between two groups. Differences were considered significant when p<0.05.

Author contributions
Huan Tuo and Feng Shu: performed all experiments. Sha She, Min Yang, Xiao Qin Zou, Juan Huang performed some experiments. Hong Ren, Huai Dong Hu and Peng Hu gave some critical suggestions to this project. Huan Tuo and Feng Shu analyzed data and wrote the manuscript. Shi Fang Peng Yi Huan Yang did conception and design and final approval of manuscript. Yi Xuan Yang made financial support, and final approval of manuscript.

CONFLICTS OF INTEREST
The authors declare that no conflicts of interest exist.