For real-time quantitative PCR detection, 12 pairs of human gastric tumor tissues and their adjacent normal gastric tissues were obtained within 30 min after surgical resection at Renji Hospital, Shanghai Jiaotong University in 2011. Twelve of these cases were chosen to extract proteins for western blot analysis. These samples were frozen immediately in liquid nitrogen for 20 sec and then stored at −80ºC until use. For immunohistochemical examinations, formalin-fixed paraffin-embedded (FFPE) samples were obtained from 64 patients who underwent surgery for gastric carcinoma. The specimens consisting of 30 normal gastric tissues, 44 chronic atrophic gastritis without intestinal metaplasia, 34 intestinal metaplasia, 36 dysplasia were obtained by biopsy during July 2009 and June 2010 at Renji Hospital. All of the patients provided informed consent, and the study protocol was approved by the Ethics Committee of Renji Hospital. An experienced gastrointestinal pathologist (X.-Y. Chen) evaluated the tumor stage and grade by microscopic examination of the samples following laser-capture microdissection.

Fresh-frozen samples were embedded in optimal cutting temperature compound (Sakura Finetek, Torrance, CA, USA) and then serially cut into 8-μm sections in a cryostat at −20ºC. The first section was mounted on ordinary glass slide and stained with H&E for diagnosis, and the other was mounted on an ethylene vinyl acetate (EVA) membrane, followed by incubation in 70% ethanol (1 min), ddH₂O (10 sec), Mayer’s hematoxylin (30 sec), twice ddH₂O (20 sec), eosin (5 sec), 95% ethanol (30 sec) and 100% ethanol (1 min). After air-drying, the sections were subjected to laser-capture using an LMD system (Leica Microsystems, Wetzlar, Germany), and the cells of interest were selectively collected following the manufacturer’s recommendations within 40 min. In order to harvest a sufficient amount of cells, 8 h was required for real-time PCR and 15 h for western blot analysis, respectively. Each dissection was determined to be >95% homogeneous by review of the histologic sections. The captured malignant and normal cells were preserved in 0.5-ml microcentrifuge tubes containing 200 μl TRIzol reagent for isolating total RNA or containing nothing for extracting protein and then stored at −80ºC until use.

RNA was isolated from the captured cells using the Optimum FFPE RNA Isolation kit (Ambion, Austin, TX, USA) following the recommended manufacturer’s instructions. Chloroform (40 μl) was added to each tube, and the tubes were shaken violently by hand for 15 sec before incubation at room temperature for 5 min. After centrifugation at 12,000 rpm for 15 min at 4ºC, the aqueous layer was transferred to a 1.5-ml Eppendorf tube. Glycogen (1 μl) (10 μg/μl) carrier and 100 μl isopropanol were added successively and precipitation was carried out at room temperature for 10 min followed by centrifugation at 12,000 rpm for 15 min at 4ºC. The RNA pellet was washed in 500 ml 75% ethanol, dried and redissolved in 20 μl diethylpyrocarbonate-treated RNase-free water before storing at −80ºC. RNA samples were treated with DNase (Roche Diagnostics Corp., Indianapolis, IN, USA), and then reverse transcription was performed using 4 μl of 5× PrimeScipt™, 1 μl of Oligo(dT) primer (50 μM), 1 μl of random 6 mers (100 μM), 1 μl of PrimeScript™ RT Enzyme Mix I, × μl (1 μg) of total RNA and (13-x) μl of RNase-free H₂O in a total volume of 20 μl, followed by incubation at 37ºC for 15 min (first-strand synthesis) and at 85ºC for 5 sec (inactivation of the reverse transcriptase) then the samples were maintained at 4ºC until use.

Real-time quantitative PCR was performed using the ABI PRISM 7300 sequence detection system (ABI-7300H; Applied Biosystems, Foster City, CA, USA) in a 10-μl reaction system including 5 μl of SYBR^® Premix Ex Taq™ II (2X) (Takara Corp.) 0.4 μl of PCR forward primer (10 μM), 0.4 μl of PCR reverse primer (10 μM), 0.2 μl of ROX Reference Dye (50X), 1 μl of mina53 cDNA or c-myc cDNA and 3 μl of dH₂O. All the reactions were performed in triplicate. The sequence of primer pairs were: sense, GGG ACA CAA CAT TGG GTA TCA TCA and antisense, AAC ATG GGC AAT TCA GGC AGA for mina53; sense, CCA CAG CAA ACC TCC TCA CAG and antisense, GCA GGA TAG TCC TTC CGA GTG for c-myc; and sense, GTG AAG GTC GGA GTC AAC G and antisense, TGA GGT CAA TGA AGG GGT C for GAPDH (reference gene). The reactions were carried out at 95ºC for 30 sec initially followed by 40 cycles of 95ºC for 5 sec, 60ºC for 31 sec and 72ºC for 30 sec. Reaction products were identified by agarose gel electrophoresis and GoldView staining. The threshold cycle (Ct) of every target gene for each sample was assessed using the 2^−ΔΔCt method [ΔΔCt = (tumor gene Ct - GAPDH Ct) − (normal epithelial gene Ct - GAPDH Ct)]. A fold-change ≥2 was determined to be high expression.

Cells from 12 pairs of samples were lysed in a mixed buffer (9.5 mol/l urea, 65 mmol/l DTT, 4% CHAPS, 0.2% IPG buffer) for 1 h in an ice-bath and then centrifuged at 12,000 rpm for 1 h at 4ºC. Sample proteins (20 μg) were subjected to 8% sodium dodecylsulfate-polyacrylamide gel electrophoresis followed by electrotransfer onto a polyvinylidene difluoride microporous membrane (Millipore, Bedford, MA, USA) and then blocked with 5% skim milk/0.01% Tween for 2 h at room temperature. After treatment with rabbit anti-Mina53 polyclonal antibody at a dilution of 1:2,500 (Abcam Biotechnology), rabbit anti-c-Myc monoclonal antibody (Bioget Technology) at a dilution of 1:1,000 or HRP-conjugated rabbit anti-GAPDH monoclonal antibody (Shanghai Kangcheng Biotechnology Co., Ltd., Shanghai, China) at a dilution of 1:3,000 at 4ºC overnight and incubated with HRP-conjugated secondary antibodies except for GAPDH at room temperature for 1 h, blots were then detected with enhanced chemiluminescence for 1 min (Pierce Biotechnology).

Routinely processed FFPE serial sections (4 μm) were mounted on coated slides and deparaffinized in xylene and graded alcohol. After pretreatment with 3% H₂O₂, deparaffinized 4-μm tissue sections were autoclaved for Mina53. After treatment with 5% non-immune goat serum, the sections were incubated overnight at 4ºC with the anti-Mina53 polyclonal antibody at a dilution of 1:500 (Abcam Biotechnology), followed by peroxidase-labeled goat anti-mouse or goat anti-rabbit IgG Fab’. Color was developed with 3,3-diaminobenzidine before the light counterstaining with hematoxylin. The sections were then air-dried, coverslipped and observed with an Olympus BX51 microscope (Olympus Optical Co., Ltd., Tokyo, Japan).

The positive staining intensity of cells was scored as follows: 0, no staining; 1, shallow brown; 2, brown; 3, dark brown. The percentage of positive cells was divided into 4 levels: 0, ≤10% positive cells; 1, 11–50% positive cells; 2, 51–75% positive cells; 3, >75% positive cells. The total score for the staining of samples was calculate by multiplication of the intensitiy and percentage scores: 0, (−); 1–2, (+); 3–4, (++); 5–9, (+++). Each score ≥3–4 (++) was considered to be high expression.

The comparison of the mRNA and proteins between the malignant cells and the adjacent non-malignant cells was tested using Wilcoxon signed rank test, and the relationship between the mRNA or protein of mina53 and c-Myc was checked by Pearson correlation analysis. The staining index ratio of Mina53 was compared between different tissues using ANOVA analysis of variance. The correlation between Mina53 expression and Lauren’s type was examined using χ² test, and the same method was carried out between Mina53 expression and the various clinicopathological factors in IGC and DGC, respectively. Crude survival curves were calculated using the Kaplan-Meier method. P-values <0.05 were assigned to indicate statistically significant results. All statistical analyses were carried out using SPSS 16.0 (SPSS, Inc., Chicago, IL, USA).

When compared with the adjacent normal epithelial tissues, mina53 and c-myc mRNA expression in the malignant tissue was highly increased (P<0.05 and P<0.01) as determined using real-time quantitative PCR. Mina53 expression was higher (fold-change ≥2) in 9 cases and c-Myc expression was higher (fold-change ≥2) in 12 cases (Fig. 1). A significant positive correlation between mina53 mRNA and c-myc mRNA in the gastric carcinoma was noted (r=0.58, P<0.05).

The intensity of the band for Mina53 in the malignant tissue was increased when comparing to that in the adjacent normal epithelial samples in all 12 cases (P<0.01) (Fig. 2), and the expression level of c-Myc was found to be higher in malignant cells when compared to that in the adjacent normal epithelial tissues (P<0.05). Nevertheless, the c-Myc expression level was observed to be lower in the malignant than that in the matched non-malignant tissues in 3 (25%) cases. The Mina53 expression was found to be significantly correlated with that of c-Myc (r=0.876, P<0.01).

The percentage of cases with overexpression of Mina53 was determined in normal gastric tissues (0%), chronic atrophic gastritis without intestinal metaplasia (9.1%), intestinal metaplasia (29.4%), dysplasia (41.7%) and IGC tissues (50%). The frequency of the overexpression of Mina53 was progressively increased (Table I), with a statistically significant difference (P<0.01). Mina53 expression was observed to be mainly localized in the nucleus and partly in the cytoplasm. Only a few cells in the neck of the glands eliminating mature foveola gastricae and proper gastric glands were detected to express Mina53 with low intensity in normal gastric tissues. The expression of Mina53 was increased in chronic atrophic gastritis without intestinal metaplasia which was still mainly located in the neck of the glands. In chronic atrophic gastritis with intestinal metaplasia, the expression of Mina53 was observed in intestinal metaplasia excluding proper gastric glands (Fig. 3).

No relationship was observed between the percentage of cases with Mina53 overexpression and patient gender and tumor site in both types of cancers. High expression of Mina53 was found to be correlated to age (P<0.05), tumor diameter (P<0.05), depth of invasion (P<0.01), lymph node metastasis (P<0.05), distant metastasis (P<0.05) and TNM staging (P<0.05) in IGC cases, while high expression of Mina53 was found to be correlated only with lymph node metastasis (P<0.01) in DGC cases. The Mina53 expression level was higher in the groups with age ≤63 years, tumor diameter >5 cm, depth of invasion (T3+T4), lymph node metastasis, distant metastasis and TNM stage (II, III, IV) in the IGC cases (Table II).

Fifty patients successfully followed up were divided into two groups according to low or high Mina53 expression. Crude survival curves were estimated for each group using the Kaplan-Meier method. The survival rate was lower in patients with tumors with high mina53 expression than in patients with tumors with low mina53 expression (P=0.007; Fig. 4).