A total of 124 gastric tissues, 62 matched normal and carcinoma pairs, were obtained from patients who underwent surgery at Miyagi Cancer Center (Natori, Japan), between 2007 and 2013. All samples were immediately frozen after resection in liquid nitrogen and stored at −80°C or fixed in 10% buffered formalin and embedded in paraffin wax.

The gastric cancers were histopathologically classified as the intestinal type and diffuse type according to the classification of the World Health Organization, and were additionally classified according to the pathologic tumor-node-metastasis (TNM) Classification (15). No patients received chemotherapy or radiotherapy before surgery. For statistical analysis, overall survival was defined by death from any cause, and Kaplan-Meier survival curves were used.

The gastric cancer cell lines MKN74 (intestinal type), GCIY (diffuse type) and HGC-27 (diffuse type) were obtained from RIKEN BioResource Center (Tsukuba, Japan). MKN74 was maintained in RPMI-1640 (Wako Pure Chemical Industries, Osaka, Japan) and GCIY and HGC-27 were maintained in DMEM (Wako Pure Chemical Industries), containing 10% inactivated FBS (EuroClone, Milan, Italy) with 100 U/ml penicillin and 100 μg/ml streptomycin (Nacalai Tesque, Kyoto, Japan) and were cultured in a humidified 5% CO₂ incubator at 37°C.

Total RNA was extracted from frozen samples and cell lines using RNeasy Mini kit (Qiagen, Tokyo, Japan) according to the manufacturer’s protocol. First-strand cDNAs from all samples were synthesized from 1.0 μg of total RNA by PrimeScript^® 1st strand cDNA Synthesis kit (Takara Bio, Shiga, Japan) following the manufacturer’s protocol. The expression of PLXND1 and SEMA3E was quantified by LightCycler Brilliant SYBR Green qRT-PCR kit (Roche Applied Science, IN, USA) according to the manufacturer’s protocol with the specific primer sets (Table I). The levels of PLXND1 and SEMA3E expression in each sample were normalized to the respective GAPDH expression levels. The specificity of each PCR reaction was confirmed by melting curve analyses.

To assess the phosphorylation of Erk in MKN74 by SEMA3E, MKN74 cells were plated in the culture medium without FBS overnight. The culture medium was aspirated from the dish and cells were washed using PBS. Culture medium with or without recombinant SEMA3E was injected. Five minutes post-injection, the cells were harvested and western blot analysis was performed using antibodies of α-tubulin, Erk and phosphorylated Erk.

To knockdown SEMA3E in GCIY and HGC-27, we used Knockout™ RNAi systems (Clontech Laboratories, Mountain View, CA, USA) according to the manufacturer’s protocol. We designed seven shRNA sequences targeting SEMA3E according to a previous study (11). After annealing of the complementary shRNA oligonucleotides, we ligated those oligonucleotides into pSIREN vector (sh1 and 2). Then, we transfected Platinum-A packaging cell lines (Provided by Professor Kitamura) with shSEMA3E or pSIREN Vector (control) to produce recombinant retroviruses. Stably infected GCIY and HGC-27 cells with the recombinant retrovi-ruses were selected with puromycin. The cells transduced with the sh1 and -2 vector were selected for further study since these cells showed effective reduction of SEMA3E by qRT-PCR. The sequences of shRNA targeted SEMA3E are shown in Table II. All the experiments in vitro were performed in triplicate, repeated three times and representative data are shown.

Human SEMA3E cDNA (Kazusa DNA Research Institute, Chiba, Japan) was amplified by PCR and inserted into the pcDNA4.0/3xHA-HisA vector (pcDNA4.0/3xHA-HisA-SEMA3E). Then, human SEMA3E cDNA with HA tag was amplified by PCR and inserted into the pBabe puro vector (pBabe-SEMA3E-HA). Recombinant retroviruses were produced with Platinum-A packaged cell lines. Briefly, Plat-A cells were transfected with pBabe-SEMA3E-HA or pBabe-puro Vector (EV). FuGENE-6 (Roche Applied Science) and Opti-MEM I (Gibco/Life Technologies Co.) were added following the manufacturer’s protocol. Forty-eight hours after transfection, the retrovirus-containing supernatant was collected and passed through a 0.45-μm filter. MKN74 cells were infected with the recombinant retroviruses and then selected with puromycin.

Antibodies used in this study were obtained from the following sources: α-tubulin (Santa Cruz Biotechnology, CA, USA): Hemagglutinin (HA) (Roche); Erk, and phosphorylated Erk (Cell Signaling Technology, MA, USA). Recombinant SEMA3E was purchased from R&D Systems (R&D Systems, MN, USA).

Cells were harvested in RIPA buffer supplemented with protease and phosphatase inhibitors (Roche). The lysates were pre-cleared by centrifugation (10,000 × g) for 15 min at 4°C, and supernatants were subjected to 5–20% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto polyvinyliden difluoride membranes. After blocking with 5% non-fat milk in Tris-buffered saline (TBS) containing 0.1% Tween-20, the membranes were incubated overnight at 4°C with the primary antibody. After washing with TBS buffer, they were incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies. Reactive bands were detected using SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher Scientific Inc., MA, USA).

The cell growth rate was evaluated using Cell Counting Kit-8 (Dojindo, Kumamoto, Japan). Cells (3×10³) of GCIY, HGC-27 and MKN74 per well were seeded in 96-well plates in normal cell growth medium. After incubation for 24, 48 or 72 h, Cell Counting Kit-8 solution (10 μl) was added and the cells were further incubated for 2 h at 37°C. Then, measurements of absorbance at 450 nm by VersaMax (Molecular Devices, CA, USA) were measured after 24-, 48- and 72-h incubation. The index evaluated at 48 and 72 h was normalized to that at 24 h.

Soft agar assays were conducted in 6-well plates. The base layer of each well consisted of 1 ml with final concentrations of 1X medium (DMEM or RPMI plus 10% FBS) and 0.6% low melting point agarose. Plates were cooled at room temperature until solid, at which point a 1 ml growth agar layer was poured, consisting of 5×10⁴ cells of GCIY and 1×10⁴ cells of MKN74 suspended in 1X media and 0.3% low melting point agarose. Plates were again cooled at room temperature until the growth layer congealed. A further 1 ml of 1X medium without agarose was added on top of the growth layer on day 0 and 0.5 ml was applied once a week. Cells were allowed to grow at 37°C for 4 weeks and total colonies were counted.

Six-week-old female NOD/Shi-scid-IL2Rγ^null (NOG) mice were purchased from Central Institute for Experimental Animals (CIEA, Kawasaki, Japan) and used in this experiment. They were allowed to acclimate for a week before the experiments. They were housed in a clean room of the animal care facility at Miyagi Cancer Research Center and kept under standard temperature, humidity and timed lighting conditions and provided mouse chow/water ad libitum. The approval of the Animal Welfare Committee at the Institute was obtained for all studies.

To investigate the effect of SEMA3E expression in tumor formation and growth in vivo, 1×10⁵ cells of GCIY-control (n=6) or GCIY-sh1 (n=6) were injected subcutaneously with 100 μl PBS into NOG mice. Then we measured the volume of the subcutaneous tumors formed once a week for 8 weeks. The tumor volume was calculated by the formula: [(long diameter] × (short diameter)²] × 1/2.

To assess the effect of SEMA3E expression on peritoneal metastasis, we injected 1.0×10⁶ cells of GCIY-control (n=6) or GCIY-sh1 (n=6) with 200 μl PBS into the peritoneum of mice. After 5.5 weeks, the mice were sacrificed, and the presence of nodules and ascites in the peritoneal cavity was investigated macroscopically.

The correlation of PLXND1 and SEMA3E expression with the patient’s clinicopathological variables was analyzed by Fisher’s exact test. Statistical significance of differences between 2 groups was determined using Student’s t-test or Wilcoxon rank sum test and p-value (<0.05) was regarded as statistically significant. Statistical difference between 3 or 4 groups was evaluated using Tukey’s test and a p-value <0.05 was regarded as statistically significant. All statistical analyses were performed using JMP^® 10 (SAS Institute Inc., Cary, NC, USA).

The SEMA3E and PLXND1 expressions of cancerous and adjacent non-cancerous tissues from the gastric cancer patients were examined by qRT-PCR and normalized by the GAPDH expression levels, and we evaluated the ratios of cancer/non-cancer expression levels of SEMA3E and PLXND1. The human gastric cancer samples were obtained from patients with tumors localized to the stomach (N0) or with any lymph node involvement or distant metastases at diagnosis (N+M+). The mean expression ratio of PLXND1 was 1.27±1.73 and 1.50±2.40 (cancer/non-cancer of PLXND1 ± standard deviation) in N0 (n=21) and N+M+ (n=41) patients, respectively. The mean expression ratio of SEMA3E was 0.52±0.72 and 2.73±6.05 (cancer/non-cancer of SEMA3E ± standard deviation) in N0 (n=21) and N+M+ (n=41) patients, respectively. While there were no significant differences in the PLXND1 expression levels between N0 and N+M+ patients (p=0.9171), the SEMA3E expression levels in N+M+ patients were significantly higher than those in N0 patients (p=0.0496) (Fig. 1A).

Moreover, the human gastric cancer tissues were classified according to the SEMA3E expression ratio into a high-SEMA3E group (cancer/non-cancer ≥1.5, n=15) and low-SEMA3E group (cancer/non-cancer <1.5, n=47) (Fig. 1B). The association between the SEMA3E expression ratio and the clinicopathological features is summarized in Table III. A significant association was found between the SEMA3E expression ratio and the diffuse type gastric cancer. The PLXND1 expression ratio was higher in the high-SEMA3E group than in the low-SEMA3E group, but there was no significant difference.

In addition, in the intestinal type gastric cancer patients, the high-SEMA3E group (n=5) showed shorter survival than the low-SEMA3E (n=32) group (p=0.0017, Fig. 1C).

At first, the PLXND1 and SEMA3E expression levels of several gastric cancer cell lines were examined by qRT-PCR and normalized by the GAPDH expression level (Fig. 2A). Since PLXND1 expression was not associated with the biological aggressiveness of gastric cancer, we examined whether SEMA3E could activate Erk, which is the one of the representative growth promoting oncogenic signals, in gastric cancer cells with low level expression of PLXND1. As shown in Fig. 2B, SEMA3E could phosphorylate Erk in MKN74 cells, which endogenously express very low level of PLXND1, depending on the concentration of recombinant SEMA3E.

To assess the functional role of SEMA3E in gastric cancer, we generated stably shSEMA3E-expressing GCIY and HGC-27 cells, which express very high levels of endogenous PLXND1 as well as SEMA3E. The qRT-PCR revealed that the SEMA3E expression levels were effectively reduced in the knockdown cells (Fig. 2C).

MKN74, a gastric cancer cell line derived from intestinal adenocarcinoma, expressed low levels of PLXND1 and SEMA3E. Therefore, we generated stably SEMA3E-overexpressing MKN74 cells (MKN74-SEMA3E-HA). The qRT-PCR and western blot analysis revealed that the SEMA3E expression levels were effectively increased in MKN-SEMA3E-HA (Fig. 2D). Consistent with activation of Erk by SEMA3E in MKN74, MKN74-SEMA3E-HA cells phosphorylated Erk more than MKN74-EV (Fig. 2B).

To assess the functional role of SEMA3E in gastric cancer development, we utilized Cell Counting Kit-8 assays. The results of these assays showed that cell growth was significantly reduced in SEMA3E knockdown GCIY and HGC-27 cells (sh1 and sh2) (Fig. 3A and B). In addition, the cellular proliferation of MKN74 cells was significantly increased in a SEMA3E-dependent manner (Fig. 3C) and MKN74-SEMA3E-HA significantly proliferated more than MKN74-EV (Fig. 3D).

To evaluate the function of SEMA3E in the anchorage-independent cell growth, we utilized a soft agar assay. GCIY-sh1 cells formed a smaller numbers of colonies than respective control cells (p=0.001, Fig. 4A). Consistent with this finding, MKN74-SEMA3E-HA cells formed larger numbers of colonies than MKN74-EV in anchorage-independent cell growth (p=0.007, Fig. 4B).

The effects of SEMA3E expression on gastric cancer growth and peritoneal dissemination in vivo. We examined the functional role of SEMA3E in gastric cancer development by subcutaneous injection assay of NOG mice. The tumor growth in NOG mice was significantly impaired in SEMA3E knockdown GCIY cells (sh1) (p<0.001) (Fig. 5). Next, we examined the metastatic capacity of SEMA3E knockdown gastric cancer cells by intraperitoneal administration assay. GCIY-sh1 cells macroscopically metastasized more slowly (Fig. 5A) and in smaller numbers (Fig. 5B) of mesentery nodules in the peritoneal cavity compared to control cells (p=0.06, 0.008) (Fig. 6).