Human GC cell lines AGS, Hs746T, KATOIII, PSK4, MKN-1, MKN-28, MKN-74, NCI-N87, SNU1, SNU16, SNU216, SNU484, SNU601, SNU638, SNU668, and SNU719 were cultured in RPMI-1640 containing 10% fetal bovine serum (FBS). Cell lines YCC2, YCC3, YCC6, YCC7, YCC9, YCC11, and YCC16, generated at Yonsei University College of Medicine, were cultured in MEM containing 10% FBS. HEK293T cells were cultured in DMEM containing 10% FBS. SNU484 and NCI-N87 cells that constitutively expressed SYT11 were selected with 50 μg/ml hygromycin. All cell lines tested for mycoplasma contamination using cycleave polymerase chain reaction (PCR) mycoplasma detection kit (Takara).
We purchased SP600125 from Sigma-Aldrich (St. Louis, MO). SYT11-antisense oligonucleotide (ASO) and NC-ASO were synthesized by Integrated DNA Technologies (Coralville, IA). The ASO sequences were as follows: SYT11-ASO#7 (5′- mA*mU*A*T*G*A*C*A*G*A*G*A*C*A*C*C*T*mG*mG-3′), SYT11-ASO#8 (5′-mU*mU*G*G*C*A*A*T*G*C*G*C*T*T*T*C*T*mG*mC-3′), and NC-ASO (5′-mC*mC*T*A*C*G*C*C*A*C*C*A*A*T*T*T*C*mG*Mu-3′).
We purchased pCMV3-HA, pCMV3-HA-SYT11, pCMV3-GFP-SYT11, pCMV3-MYC-JNK1, and pCMV3-MYC-JNK2 from Sino Biological (Wayne, PA). Mutations (T1835 and Y185F) in JNK1 were generated using the EZchange site-direct mutagenesis kit (Enzynomics, South Korea). The deletion mutants of GFP-SYT11 were generated using the EZchange site-direct mutagenesis kit. pBluscript-MKK7 was obtained from the Korean UniGene Information. MKK7 was amplified using PCR, then cloned into the pcDNA3.1-Myc or pcDNA3.1-HA plasmid between the EcoRI/BamHI or the EcoRI/XhoI sites, respectively.
We performed a retrospective review of a GC cohort database prospectively maintained at Yonsei University College of Medicine (Seoul, South Korea) to identify all patients with gastric adenocarcinoma who underwent curative D2 gastrectomy between 2000 2010 . We obtained demographic and clinicopathologic information and tumor tissue samples from 527-patients. This study was approved by the institutional review board of Severance Hospital (Seoul, Korea; 2015–3104-001). GC samples microarray data were available at the National Center for Biotechnology Information Database of GEO datasets under the data series accession numbers GSE13861 and GSE84437 .
Lentiviral shRNA library construction
Patients with GC were classified by molecular subtypes based on microarray. We selected genes demonstrating an increase beyond twofold in the stem-like molecular subtype compared to the intestinal molecular subtype. Among them, 118 genes exhibited a raw data average of > 500 in the microarray and p < 0.05 in the KM analysis. Finally, arrayed 583-shRNA modules were constructed for 118-genes (MISSION shRNA library; Sigma-Aldrich).
In vivo and in vitro shRNA screening
We performed shRNA library infection as described previously . Briefly, the shRNA library was packed into a lentivirus via HEK293T cells using Lipofectamine 2000, then transduced into SNU484 cells with a multiplicity of infection of 0.3 and a fold representation of 500. After 48 h of transduction, SNU484 cells were subjected to puromycin selection for 72 h. Transduced SNU484 cells were preserved as a reference sample. For in vivo screening, 5 × 106 cells were injected into mice (n = 15). Tumors were extracted 4 to 6 weeks after injection, when their size was about 400 ± 200 mm3. For the in vitro screening, cells were harvested weekly for 6 weeks. The shRNA insert barcodes were amplified from the genomic DNA of tissues and cells using Pfu PCR premix (Bioneer, Daejeon, Korea), then sequenced using the Illumina Hi-Seq 2500 system (Illumina). We analyzed the barcode-seq data using the barcode sequence alignment and statistical analysis (Barcas) developed by Kim SY .
Small interfering RNA (siRNA) -mediated gene knockdown
Gene knockdown was performed by introducing siRNA into the target gene using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. The siRNA sequences were, as follows: siScramble 5′-CCUACGCCACCAAUUUCGU (dTdT)-3′, siSYT11#4 5′-CAUCAAAGUGCGGAGAGACAA(dTdT)-3′, siSYT11#5 5′-AUC CUUCCUGACAAACGGCAU(dTdT)-3′, siSYT11#6 5′-CCUGCUAAGCCGAGACAAA (dTdT)-3′, siSYT11#7 5′-CCAGGUGUCUCUGUCAUAU(dTdT)-3′, siSYT11#8 5′-GCA GAAAGCGCAUUGCCAA(dTdT)-3′. siRNAs for siANGPTL2 (23452–1), siTHBS4 (7060–1), siJAM3 (83700–1), siVimentin (7431–1), and siMKK7 (5609–1) were purchased from Bioneer.
All animal experiments were approved by the bioethics committee of the Korea Research Institute of Bioscience and Biotechnology. We performed in vivo xenografts as described previously . We injected lentiviral shSYT11 vector infected SNU484 cells (5 × 106) subcutaneously into 5-week-old female BALB/c nude mice. To test SYT11-ASO antitumor efficacy, we injected MKN1 cells (1 × 107) subcutaneously into 5-week-old female BALB/c nude mice. After 1 week, SYT11-ASO [10 mg/kg in 100 μl phosphate-buffered saline (PBS)] was administered via intraperitoneal injection 5 times a week. For the liver metastasis tail vein injection assay, we injected shControl- or shSYT11-expressing lentivirus-infected SNU484 cells (2 × 106), suspended in 100 μl of PBS, into the tail vein of BALB/c nude mice (4 mice per group). After 16 weeks, we removed and fixed the mouse livers. Tumor metastasis to the liver was assessed with Hematoxylin and eosin (H&E) staining. For metastasis quantitation, random field photos were obtained at a magnification of 40× (three fields per mouse) and analyzed using the NIH ImageJ software (version 1.48).
Western blot analysis
We lysed the cells with RIPA buffer (Millipore, Billerica, MA) containing a protease inhibitor cocktail (Roche, Basel, Switzerland), then quantified the lysates with a protein assay kit (Bio-Rad, Hercules, CA). We used sodium dodecyl sulfate-polyacrylamide gel electrophoresis to separate the cell lysates. Appropriate antibodies were used for protein identification (Supplementary Table 1).
Reverse transcription-PCR (RT-PCR) and quantitative real-time PCR (qPCR)
Total RNA was isolated using the TRIzol reagent (Invitrogen, Carlsbad, CA). cDNA was synthesized using the TOPscript™ RT DryMIX (Enzynomics, Daejeon, Korea). RT-PCR was performed using the Dr. Taq MasterMix (Doctor Protein, Daejeon, Korea). qPCR was performed using a SYBR Green master mix kit (Qiagen, Valencia, CA). The following primers were used: RPL13A (5′-CTGGACCGTCTCAAGGTGTT-3′ and 5′-TGGTACTTCCAGCCAACCTC-3′), JAM3  (5′-CTGCTGTTCACAAGGACGAC-3′ and 5′-CAGATGCCCAACGTGATCAG-3′). The SYT11 (P281379), GAPDH (P267613), ANGPTL2 (P302397), THBS4 (P266439), and Vimentin (P324997) primers were purchased from Bioneer.
Live-cell assay for cell proliferation and apoptosis
Cell confluence-based proliferation rates were measured with live-cell imaging (IncuCyte ZOOM System, Essen BioScience, Ann Arbor, MI) as described previously . To analyze apoptosis, we performed kinetic caspase-3/7 measurements using the CellPlayer reagent (Essen BioScience) as described previously . We imaged the cell frames incubated in 96-well plates at 2 h intervals from four separate regions per well using a 10× objective lens. The cultures were maintained in a 37 °C incubator.
Sulforhodamine B (SRB) and invasion assays
We measured cell viability using the SRB assay as previously described . The cells were fixed with 10% formalin and stained with 0.4% SRB. After 10 min, the cells were washed with 0.01 M acetic acid. The protein-bound dye was dissolved in 10 mM Tris, and its optical density was measured using a spectrophotometer at 540 nm.
For the invasion assays, we used chambers with 8.0-μm-pore PET membrane in 24-well cell culture inserts (BD Biosciences, San Jose, CA). We seeded the cells into the upper part of each chamber with Matrigel coating, whereas we filled the lower compartments with the above-mentioned medium. We then allowed the cells to invade, subsequently fixed them with 10% formalin, and stained with 0.4% SRB.
Fluorescein isothiocyanate (FITC)-Annexin V/ propidium iodid (PI) double-staining
We performed the FITC-Annexin V/PI double-staining analysis according to the manufacturer’s protocol (BD Biosciences). We treated the cells with ASO, then washed twice with pre-chilled PBS. We stained the treated cells with FITC-Annexin V staining buffer and PI solution for 15 min, then analyzed with a FACSCalibur Flow Cytometer (BD Biosciences).
Tissue array blocks of human GC and normal tissues were supplied by US Biomax (Rockville, MD). IHC was performed as previously described . We incubated the slides with anti-SYT11 antibodies. After washing with PBS, we incubated the slides with biotinylated anti-rabbit IgG (Vector Laboratories, Burlingame, CA) and avidin-biotin-peroxidase (Vector Laboratories), then visualized using diaminobenzidine tetrahydrochloride (Vector Laboratories). We counterstained the sections with hematoxylin.
Immunofluorescence staining was performed as described previously . Cells were incubated with anti-SYT11, anti-MKK7, and anti-JNK antibody at 4 °C O/N and then incubated with secondary fluorescent (Alexa-546 or FITC) antibodies for 1 h. Subsequently, the cells were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) and analyzed under a confocal microscope (LSM 5 LIVE DuoScan, Carl Zeiss, Stuttgart, Germany).
We induced three-dimensional spheroid cultures as described previously [23, 28].
To compare the gene expression differences between the patient subgroups, we performed two-sample t-tests. We calculated Pearson’s correlation coefficients to evaluate genetic associations. To divide the patients into two single gene expression subgroup, we obtained an optimal gene expression cutoff from the ROC analysis, determining the best cutoff by the expression with the highest multiply of sensitivity and specificity. Statistical analyses were performed using MedCalc version 18.11.6 (MedCalc Software, Ostend, Belgium). We used the Kaplan-Meier method to calculate the time before death and measured the difference between the times using the log-rank test.