Tissue samples were obtained from 352 patients with GC who underwent surgical resection at the Department of Gastric Cancer and Soft Tissue Sarcoma, Fudan University Shanghai Cancer Center from April 2010 to April 2011. No preoperative chemotherapy or radiotherapy had been performed in any of these cases. In the present study, there were 275 men and 77 women with a mean age of 58 years (range, 32–88 years). The clinical data of patients were obtained from the medical record of Fudan University Shanghai Cancer Center (Shanghai, China). Ethical approval for human subjects was obtained from the Research Ethics Committee of Fudan University Shanghai Cancer Center, and written informed consent was obtained from all patients prior to enrollment in the present study. One pair of tissue samples was obtained from each of the 352 patients. All patients were regularly followed-up by telephone. The duration of follow-up was defined as the interval from the date of the treatment to the date of mortality or the final follow-up, with the final follow-up being the 15th of December 2014. The clinicopathological characteristics of patients in the TMA are summarized in Table I. Tumor size was defined as the maximum diameter of the tumor. Tumor-node-metastasis (TNM) classification of gastric carcinoma was based on the 7th edition of the American Joint Committee on Cancer staging system (18). Borrman type was classified as previously described (19). Histological grade was classified according to the Japanese Classification of Gastric Carcinoma as well, moderately and poorly differentiated (20). Preoperative carcinoembryonic antigen (CEA) levels were obtained using an ELISA kit according to the manufacturer's protocol (cat no. ab99992; Abcam, Cambridge, UK). Briefly, the assay employed anti-human CEA antibody coated onto a 96-well plate. Standard or serum samples were pipetted into the wells and CEA present in samples was bound to the wells by the immobilized antibody. The wells were washed and biotinylated anti-human CEA antibody was added. Following the washing of unbound biotinylated antibody, horseradish peroxidase-conjugated streptavidin was pipetted into the wells. The wells were washed again, tetramethylbenzidine substrate solution was added and color developed in proportion to the amount of bound CEA. Absorbance value was measured at 450 nm. CEA concentration in serum was achieved according to standard curves.

TMAs were constructed as described previously (21). In brief, identical 1.5 mm diameter cylinders from the center of the tumor were included from each case. Cylinders from the donor blocks were transferred to the recipient paraffin block at defined array positions. Consecutive 4 mm sections were placed on 3-aminopropyltriethoxysilane coated slides (in collaboration with Shanghai Biochip Company, Shanghai, China). Hematoxylin and eosin-stained slides from each tissue block were reviewed by a senior consultant pathologist, and performed according to a previously described protocol (22).

IHC was performed as previously described (23). Sections (4 µm) of primary gastric adenocarcinoma and the adjacent normal gastric tissues were subjected to immunohistochemical staining as follows: Formalin-fixed and paraffin-embedded tissue sections were deparaffinized in xylene, rehydrated in a graded alcohol series and washed with PBS. Subsequently, the tissue sections were immersed in 10 mmol/l citrate buffer (pH 6.0) and heated in a microwave for 30 min. Following cooling to room temperature, endogenous peroxidase was blocked by incubation with 3% H₂O₂ in methanol (36°C, 60 min). Nonspecific binding was blocked by incubating the tissue sections with 1% bovine serum albumin (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) in a humid chamber for 60 min at 4°C. Incubation with the primary antibodies was subsequently performed overnight at 4°C using mouse anti-human MTBP (dilution, 1:200; cat no. sc-99047; Santa Cruz Biotechnology, Inc., Dallas, TX, USA). The negative controls were treated identically but with the primary antibodies omitted. Subsequently, incubation with suitable secondary antibodies (cat no. pk-6100; Vector Laboratories, Burlingame, CA, USA) at room temperature. Finally, diaminobenzidine (Vector Laboratories) was used for signal development, and the sections were counterstained with 20% hematoxylin for 8 min at room temperature. Detection was performed using a Dako EnVision system (Dako; Agilent Technologies, Inc., Santa Clara, CA, USA).

The expression level of MTBP in tumor and normal adjacent tissues was assessed by two investigators blinded to the clinical data and to the other investigator's score. The scores were determined by the proportion of positive tumor cells and the intensity of the coloring for MTBP, as described previously (9). MTBP cytoplasmic expression level was initially assigned an intensity grade: 0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining. Nuclear MTBP staining was scored positive or negative. The scores were categorized into 2 groups: Low expression (0/1 cytoplasmic staining) with no nuclear staining and high expression (2/3 cytoplasmic staining) and/or nuclear positivity.

The reagents and protocols used were described in detail previously (24). Paired tumor tissues and surrounding non-tumor tissues were ground to powder in liquid nitrogen and lysed with SDS-PAGE sample buffer. Primary antibodies against MTBP were purchased from Abcam (1:500; cat no. ab115529) and used according to the manufacturer's instructions. A monoclonal antibody against GAPDH (1:1,000; cat no. sc-69778; Santa Cruz Biotechnology, Inc.) was used as an internal control. GAPDH was used as a control. The primary antibodies were detected with horseradish peroxidase-conjugated secondary antibodies (1:1,000; cat no. sc-2351; Santa Cruz Biotechnology, Inc.); Immunoreactive bands were visualized using SuperSignal West Pico Chemiluminescent Substrate (cat no. NCI5080; Pierce; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The gray value of each band was measured and data are presented as a ratio to GAPDH (ImageJ 1.46; National Institutes of Health, Bethesda, MD, USA). Each experiment was repeated ≥3 times.

As described previously (25), total RNA was extracted from tumor and normal adjacent tissues using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). RT-qPCR analysis was performed using an ABI PRISM 7500 Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). Each well (20 µl reaction volume) contained 10 µl Power SYBR-Green PCR Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc.), 1 µl of each primer (5 µmol/l) and 1 µl template. The following primers were used: MTBP forward, 5′-TCCTGTAGTTTCGTCAGATCCT-3′ and reverse, 5′-CCGTTTCAATCGGGATACTTCA-3′. GAPDH forward, 5′-ACAGCCTCAAGATCATCAGCA-3′ and reverse, 5′-ATGAGTCCTTCCACGATACCA-3′. GAPDH was used as internal control. Quantification was performed as described previously (25).

All statistical analyses were performed using SPSS version 19.0 for Windows (IBM Corp., Armonk, NY, USA). The associations between MTBP expression level and clinicopathological characteristics were evaluated using the χ² test. Cumulative survival time was determined by the Kaplan-Meier method using GraphPad Prism software (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA), and differences in survival curves were analyzed using the log-rank test. Multivariate logistic regressions were used to assess the association between MTBP expression level and metastasis. Multivariate analysis was performed using the Cox's proportional hazards regression model on all significant characteristics evaluated for univariate analysis. P<0.05 was considered to indicate a statistically significant difference.

Representative data are presented in Fig. 1. RT-qPCR and western blot analysis were used to confirm mRNA and protein expression levels. The mean expression level of MTBP mRNA in normal tissues was significantly higher compared with in tumor tissues (P<0.0001; Fig. 1A). Furthermore, the MTBP mRNA expression level was significantly lower in 17/20 (85%) GC tissues compared with in the matched adjacent normal gastric mucosa tissues (Fig. 1B).

The difference in MTBP expression levels between tumors and normal tissues revealed at the protein level was investigated by western blotting. The representative western blotting results in six cases are presented in (Fig. 1C). The relative level of MTBP expression was normalized to the GAPDH of the same samples. Western blot analysis demonstrated that the expression level of MTBP protein was markedly decreased in 10/14 (71.4%) gastric tumor tissues compared with in the corresponding adjacent normal tissues, which was consistent with that of the RT-qPCR results. The average MTBP protein expression level in 14 GC tissues was significantly lower compared with in the matched adjacent normal gastric mucosa tissues (P<0.05) (Fig. 1D). Fig. 1E revealed that normal tissues demonstrated higher expression levels of MTBP (81%), statistical analysis revealed a lower expression level of MTBP in gastric tumors (19%) compared with in adjacent normal tissues (P<0.05).

To elucidate the biological significance of MTBP in GC, the present study examined the immunohistochemical expression levels of MTBP in GC tissues (Fig. 2). MTBP staining was mainly located in the cytoplasm and/or the nucleus of tumor cells, as presented in Table II, consistent with that of a previous study (9). As presented in Table I, there were no significant associations between the expression level of MTBP protein and age, tumor size, Borrmann type, depth of invasion, histological grade or preoperative carcinoembryonic antigen expression levels in patients with GC. However, the expression level of MTBP was significantly associated with gender, lymph node metastasis, distant metastasis and pathological tumor-node-metastasis (pTNM) stage in GC tissues.

It was revealed that low MTBP expression level was associated with higher distant metastasis and lymph node metastases rate (Fig. 3). A multivariate logistic regression analysis was performed to evaluate the independently predictive significance of MTBP expression level for distant and lymph node metastases (Table III). The results demonstrated that the tumor size [OR, 1.942; 95% confidence interval (CI), 1.031–3.659; P=0.040] and MTBP expression level (OR, 0.365; 95% CI, 0.138–0.965; P=0.042) were independently associated with distant metastasis. As described above, the multivariate logistic regression analysis revealed that MTBP expression level was significantly associated with the presence of lymph node metastasis (OR, 0.282; 95% CI, 0.161–0.494; P<0.001) (all Table III).

To investigate the prognostic value of MTBP expression level in patients with GC, overall survival (OS) analysis was performed in 352 patients with GC. The 1- and 3-year OS rates were 88.1 and 72.4%, respectively. The OS rate of the high-level expression group was significantly longer compared with that of the low-level expression group (Fig. 4).

Univariate and multivariate analyses were performed to determine the predictors for OS (Table IV). To further determine the effect of MTBP expression on OS, the present study first performed univariate analysis of traditional clinicopathological variables for prognosis. In univariate analysis, low expression level of MTBP (P=0.017), larger tumor size (P=0.001), lymph node metastasis (P=0.001), elevated preoperative CEA (P=0.047) and TNM stage (P<0.001) were revealed to be associated with a poor OS rate of patients with GC. Furthermore, to evaluate the independent impact of MTBP expression level on OS, a multivariate Cox's regression model was performed. The results demonstrated that low MTBP expression level (HR, 0.633; 95% CI, 0.417–0.961) was a poor independent prognostic factor for OS in patients with GC. In addition, tumor size (HR, 0.582; 95% CI, 0.389–0.871) and TNM stage (HR, 2.720; 95% CI, 1.395–5.305) revealed independent prognostic value in the multivariate analysis (Table IV).