Tumor samples of International Union Against Cancer stage II or III were collected from overall 45 patients who underwent curative resection for gastric cancer at Wakayama Medical University Hospital from January 2010 to December 2010. The diagnostic procedure for gastric cancer fulfilled the following criteria included for analyses in the current study: Patients with primary gastric cancer with preoperatively diagnosed by endoscopy; patients who were not administered chemotherapy before surgery; and patients with no signs of ascites, distant metastases, or bulky para-aortic lymph node metastases after physical examination and enhanced CT scan evaluation. They included 26 stage II and 19 stage III gastric cancer patients based on Tumor Node Metastasis (TNM) Classification of the International Union Against Cancer (28). The mean age of the patients was 71.3 years, and there were 30 male and 15 female subjects. The follow-up period was five years. Stage II and III patients based on TNM classification without submucosal cancer received S-1 (oral fluoropyrimidine)-based postoperative adjuvant chemotherapy. The present study was approved by the Human Ethics Review Committee of Wakayama Medical University (approval no. 1657). Informed consent was obtained in the form of opt-out on the web page of Wakayama Medical University from all patients in accordance with the guidelines of the Ethical Committee on Human Research of our institution.

Immunohistochemical analysis of hTERT expression was performed using an anti-telomerase reverse transcriptase mouse monoclonal antibody (sc-393013; Santa Cruz Biotechnology, Inc.) as described previously (29). Pretreatment was performed by autoclaving the tissues in citrate buffer (pH 6.0) for 7 min at 121°C. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol. Nonspecific binding sites were blocked with 0.25% casein in phosphate-buffered saline (PBS) containing stabilizing protein and 0.015 mol/l sodium azide. Primary antibodies were diluted in PBS, then added to the samples followed by overnight incubation at 4°C. Following two washes, the sections were incubated for 90 min at room temperature with Histofine Simple Stain MAX-PO (MULTI) (Nichirei). Finally, the reaction products were stained with a 3,3′-diaminobenzidine substrate, counterstained with hematoxylin, dehydrated with ethanol, and fixed with xylene.

For scoring assessments, the cells were counted in five separate areas of intratumoral regions under ×400 high-power magnification. The staining intensity was defined as follows: 0, no staining; 1+, weak; 2+, moderate; and 3+, strong (Fig. 1A). The predominant intensity was chosen in case of areas with different staining intensities. The quantification of positivity (0-100%) was based on an estimate of the percentage of stained cancer cells in the lesion. The final immunostaining scores were calculated by multiplying the staining intensity with the percentage of positive cells, thereby generating immunostaining scores ranging from 0 to 300 (30–32). The cutoff values of the immunostaining scores were set as the median value, as per previous reports (33–35).

Vero (African green monkey kidney normal cell line), MKN1, MKN28, MKN45, MKN74, NUGC3, NUGC4, KATOIII, and N87 (human gastric cancer cell lines) cells were obtained from the RIKEN BioResource Center. All cell lines were authenticated according to the Cell Line Verification Test Recommendations of ATCC Technical Bulletin no. 8 (2008). TMK-1 cells, a human gastric cancer cell line, were provided by Dr Eiichi Tahara (Hiroshima University, Hiroshima, Japan). All human gastric cancer cell lines were cultured in RPMI-1640 and Vero cells were cultured in Dulbecco's modified Eagle's medium both supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Inc.).

T-hTERT is a fourth-generation oncolytic HSV, which was provided to us by Dr Tomoki Todo (The University of Tokyo, Tokyo, Japan). It was constructed by deleting the α47 gene and both copies of the γ34.5 gene, with the hTERT promoter regulating the ICP6 gene expression (Fig. 2A). γ34.5 is a major determinant of HSV neurovirulence and blocks host shutoff of protein synthesis in response to viral infection. Lack of this function is likely responsible for the less efficient growth of γ34.5-mutants when compared with wild-type HSV, as observed in many tumor cell types. This double mutation confers important advantages such as minimal chance of reverting to wild-type, preferential replication in tumor cells, attenuated neurovirulence, and ganciclovir and acyclovir hypersensitivity (36). Because of the overlapping transcripts encoding ICP47 and US11, the deletion in a47 also places the late US11 gene under the control of the immediate-early a47 promoter. This alteration in US11 expression enhances the growth of g34.5-mutants by preventing protein synthesis shutoff (36). ICP6 encodes a large subunit of ribonucleotide reductase (RR), an enzyme critical for nucleotide metabolism and viral DNA synthesis in non-dividing cells but not in dividing cells. By placing the ICP6 gene under the hTERT promoter, the hTERT promoter is activated in tumor cells and expresses ICP6. T-null is an HSV-1-based oncolytic virus, constructed by deleting ICP6, α47, and both copies of γ34.5 and are not regulated by the hTERT promoter (Fig. 2A). Viral stocks were prepared by releasing the virus from infected Vero cells with heparin followed by high-speed centrifugation, as described previously (37).

T-null was used to treat gastric cancer cell lines in vitro. The cells were seeded on 6-well plates at a density of 5×10⁵ cells/well and incubated. Following a 24-h incubation, the cells were infected with T-null at 0.1 pfu/cell for 1 h and further incubated at 37°C. Cells were collected at 24 or 48 h after infection and stained with trypan blue, and the number of viable cells was counted. The survival rate was expressed as the percentage of the PBS (−)-treated control cells.

For virus yield studies, MKN1 and MKN45 cells, which are minimally sensitive to T-null, and MKN28 cells, which are moderately sensitive to T-null, were seeded on 6-well plates at a density of 5×10⁵ cells/well and incubated for 24 h. Each well was infected with either T-null or T-hTERT at 0.1 pfu/cell for 1 h and further incubated at 37°C. After 24 or 48 h of incubation, the cells were collected and stained with trypan blue, and the number of viable cells was counted. The survival rate was expressed as the percentage of the PBS (−)-treated control cells.

MKN1, MKN28, MKN45, and Vero cells were seeded on 6-well plates at a density of 5×10⁵ cells/well and incubated for 24 h. Each well was infected with either T-null or T-hTERT at 0.01 pfu/cell for 1 h and further incubated at 37°C. After a 24-h incubation, the cells were scraped and lysed by repeating the process of freezing and thawing three times. Titration of the progeny virus was measured on Vero cells via plaque assays. Each experiment was performed in triplicates.

Gastric cancer cell lines were seeded in 100 mm dishes at a density of 1×10⁶ cells/dish and incubated at 37°C. After a 24-h incubation, the cells were treated with PBS (−), T-null, or T-hTERT at 0.01 pfu/cell for 1 h, incubated further at 37°C for 24 h, and harvested. Western blot analysis was performed as described previously (38). Anti-RRM2 antibody (sc-10846; Santa Cruz Biotechnology, Inc.) or goat polyclonal anti-β-actin antibody (sc-47778, Santa Cruz Biotechnology, Inc.) were used as primary antibodies (39). Goat IgG HRP-conjugated antibody was used as the secondary antibody (HAF017, R&D Systems).

As described in a previous report (40), surgical sections of cancer tissue were collected and incubated in collagen medium for a short time (within 72 h) to evaluate the antitumor effect of oncolytic HSV. In this experiment, gastric cancer samples were collected at Wakayama Medical University Hospital from October 2016 to December 2016. They included 2 stage II and 4 stage III gastric cancer patients based on TNM Classification of the International Union Against Cancer. The mean age was 74.3 years (65–85) with 5 males and 1 female subjects. We carried out all experiments in compliance with the Declaration of Helsinki, the guidelines for ethical principles for medical research involving human subjects, and the ethical guidelines of Wakayama Medical University. This ex vivo study was also approved by the Committee of Animal Experiments and Gene Recombination (approval no. 26-31) of Wakayama Medical University. Informed consent was obtained in the written form from all patients in accordance with the guidelines of the Ethical Committee on Human Research of our institution. In general, human gastric cancer specimens collected through radical gastrectomy were incubated ex vivo on collagen gel immediately after resection. Cellmatrix type I-A collagen (Nitta Gelatin), reconstitution buffer [2.2% (w/v) NaHCO₃, 0.2 M HEPES, and 50 mM NaOH], and 10X RPMI-1640 medium were mixed at a ratio of 8:1:1 and poured into 24-well dishes (0.5 ml/well). Tissues were cut into 2 mm³ pieces and placed on collagen gel. Each well was treated with PBS (−), T-null, or T-hTERT at 0.01 pfu/cell for 1 h and incubated at 37°C for 48 h. The cell viability of gastric cancer tissues was assessed using a CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega Corp.) as per the manufacturer's instructions.

Comparison of categorial variables between hTERT expression and clinicopathological characteristics of gastric cancer patients was analyzed using the unpaired Student's t-test (for age and tumor size) and a Fisher's exact test (for sex, histological type, macroscopic type and pathological TNM classification). Multivariate analysis of overall survival was used the Cox proportional hazards regression model. The survival data were analyzed by the Kaplan-Meier method and the log-rank (Mantel-Cox) test. Quantitative data are reported as means ± standard deviation. For comparison between two groups, significant differences were determined using the unpaired Student's t-test. For comparison of multiple groups, statistical significance was determined with a one-way ANOVA followed by a Tukey's post hoc test. All analyses were performed using the SPSS statistics version 21 software (IBM Corp.). P-values <0.05 were considered statistically significant.

Immunohistochemical analyses were performed on paraffin-embedded tissues collected from 45 patients with gastric cancer. For immunohistochemistry, hTERT in gastric cancer samples was stained and observed mainly in the cytoplasm (Fig. 1A). hTERT was expressed only in cancer cells and not in normal cells (Fig. 1A). The hTERT expression scores were calculated for each sample. The median score of hTERT was 100 (range, 0–300). The binarization of the score data for this marker present in a high expression group (n=29) versus a low expression group (n=16) at the median level was performed. Patient clinicopathological characteristics are listed in Table I.

Kaplan-Meier survival curves showed the overall survival of patients with gastric cancer, characterized based on the results of hTERT expression analysis. The survival curves of the 45 patients who underwent curative R0 resection revealed a significantly poorer relapse-free survival rate in the hTERT high expression group than in the low expression group (P=0.048; Fig. 1B). Moreover, a significantly poorer prognosis was observed in the hTERT high expression group than in the low expression group (P=0.029; Fig. 1C).

Multivariate overall survival analysis was calculated using the Cox proportional hazard regression model. Multivariate analysis revealed that hTERT expression was an independent prognostic factor in patients with gastric cancer (P=0.031, hazard ratio=3.918; Table II).

After 48 h of infection with T-null at 0.1 pfu/cell, 39.0% of N87, 40.1% of NUGC3, 55.0% of TMK-1, 55.0% of MKN74, 55.1% of NUGC4, 69.1% of MKN28, and 75.6% of KATOIII cells were lysed. However, only 4.7% of MKN1 and 14.4% of MKN45 cells were lysed by T-null infection (Fig. 2B). Therefore, these results suggested that the sensitivity to T-null virus varies among human gastric cancer cell lines. Therefore, we examined the cytotoxicity of T-hTERT or T-null in MKN28 cells, which are moderately sensitive gastric cancer cell lines, and MKN45 and MKN1, which are minimally sensitive gastric cancer cell lines.

After 48 h of infection with oncolytic HSVs at 0.1 pfu/cell, the cytotoxicity of T-hTERT was found to be more than that of T-null in all cell lines. Particularly in MKN45 cells, T-hTERT showed increased cytotoxicity as compared with T-null (Fig. 2C). Therefore, we further examined the differences in viral replication between T-hTERT- and T-null-infected cells.

We compared the replication potencies of T-hTERT with those of T-null in MKN45, MKN1, MKN28, and Vero cells. The results showed that the viral titer of T-hTERT was approximately 10-fold higher than that of T-null in MKN45 cells (P<0.01). However, in MKN1 and MKN28 cells, the titers of T-hTERT were not remarkable different from those of T-null (Fig. 2D).

According to our preliminary studies, we confirmed that telomerase activity and expression of hTERT mRNA expressed in human gastric cancer cell lines (Figs. S1 and S2). RR plays an essential role in converting ribonucleoside diphosphate to 2-deoxyribonucleoside diphosphate to maintain the homeostasis of nucleotide pools (41). Human RR consists of two subunits, M1 and M2 (42). RR enzymatic activity is modulated by the expression of its M2 (RRM2) subunit (43). Therefore, we carried out a western blot analysis to examine RRM2 protein expression in the gastric cancer cell lines. Although almost all the cell lines expressed RRM2, its expression in MKN45 cells was very low (Fig. 3A). Additionally, MKN45 cells infected with T-null showed very low expression of RRM2, whereas those infected with T-hTERT showed high expression of RRM2 and its expression levels were almost 5-fold higher than those infected with T-null (Fig. 3B and C).

In terms of experimental animal protection and management, an important part of our present study was the analysis of clinical cancerous samples freshly obtained from patients with gastric cancer. As it is accepted that established cell lines differ from the initial clinical tumors, it was not completely unexpected to see a different profile of viral transfer in comparison with the cell lines. To clarify that T-hTERT remains unaffected with the heterogeneity of gastric cancer tissue, collagen gel culture consisting of gastric cancer clinical tissue samples was synthesized. To examine the effects of oncolytic HSVs in gastric cancer in vivo, human gastric adenocarcinoma specimens collected through radical gastrectomy were incubated on collagen gel immediately after resection and treated with PBS (−), T-null, or T-hTERT. After 48 h of incubation, these specimens were subjected to frozen sectioning and further examined by hematoxylin and eosin staining. In the gastric cancer specimens with oncolytic HSV infection, lysis was observed in the tumor cells (Fig. 4A and B). After 48 h of infection with T-null or T-hTERT at 0.01 pfu/cell, 36.7 and 54.9% of cells were lysed by the viruses, respectively. The T-hTERT-infected group showed significantly lower cell viability than that of the control [PBS (−)] group (P=0.029; Fig. 4C). However, as compared with T-null, T-hTERT did not show any significant cytotoxic effects (P=0.37; Fig. 4C).