Overexpression of SLC38A1 is associated with poorer prognosis in Chinese patients with gastric cancer

This study included 896 patients (median age: 61.4 years; 634 men, 262 women) with histologically confirmed gastric cancer who underwent D2/D3 curative resection at either Changhai or Changzheng Hospital in Shanghai, during a period from 2001 to 2005. Subjects who received neo-adjuvant therapy prior to the surgery were not included. Subjects with stage III disease also received a 5-fluorouracil-based chemotherapeutic regimen for 4–6 cycles. Patients presenting with neuroendocrine tumors, lymphoma or sarcoma were not included. Adjacent mucosa that was free from cancer cells was used as a healthy control.

The follow-up was conducted via phone conversation and mail in March 2010. Information with regards to survival/death was obtained in 673 cases (median survival: 59.08 months).

Demographic information and the construction of TMA blocks were described previously [12‐15]. All tissue specimens were obtained with informed consent, and the use of the human specimens was approved by Institutional Review Board at Changhai and Changzheng Hospitals.

Immunohistochemical staining was carried out on 4-μm paraffin sections after heat-mediated antigen retrieval. Samples were incubated with a human-anti-rabbit polyclonal antibody to SLC38A1 (dilution 1:200; Abcam, Cambridge, U K) overnight at 4°C. Goat anti-rabbit antibody (dilution 1:4000; Invitrogen, Carlsbad, CA, USA) was used as a secondary antibody. An immunoglobulin-negative control was used to rule out non-specific binding. As for the positive control, we referred to other studies in liver cells by Nobuo Kondoh [6] and in hilar cholangiocarcinoma by Yu WL [16], which indicated the SLC38A1 protein was stained brown and diffused in the cytoplasm. Slides were counterstained with hematoxylin.

Data (staining in cytoplasma) were acquired independently by two investigators (Guan Zhen Yu and Ying Chen) blinded to sample nature using an Olympus CX31 microscope (Olympus, Center Valley, PA, USA), and analyzed using a semi-quantitative scoring system. Staining was graded on a scale of 0–2 (0 = negative staining [no cytoplasmic staining of any tumor cells], 1 = moderate expression [cytoplasmic staining of <25% of tumor cells], and 2 = high expression [cytoplasmic staining of ≥25% of tumor cells]) as described previously [14, 17].

SH-10-TC cells were obtained from Shanghai Institute for Biological Sciences, Chinese Academy of Sciences (CAS), and grown in RPMI-1640 medium containing 10% FBS. Cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂. The experiments were carried out in the exponential phase of growth.

Four pairs of anti-SLC38A1 siRNAs and randomized cocktails of dsRNA as a control were synthesized by GenePharma (Shanghai, China). The siRNA transfection was carried out using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA).

Cell lysate containing 100 μg protein was resolved on SDS-PAGE, transferred onto PVDF membrane and incubated with an anti-SLC38A1 antibody (dilution, 1:500; Invitrogen, Carlsbad, CA, USA). Protein bands were evaluated using an Odyssey infrared fluorescent scanning system (LI-COR Biosciences, Lincoln, NE, USA) and Scion Image software (NIH, USA). GAPDH was used as an internal control.

SH-10-TC cells were seeded in 96-well plates at the day before transfection. Cells were counted at 48 h, 72 h and 96 h after transfection using a microtiter-plate colorimetric WST assay (Cell counting kit-8; Dojindo Laboratories, Kumamoto, Japan) at 450 nm. Experiments were conducted in triplicate.

Assays were performed using a modified Boyden chamber (Corning Costar, Rochester, NY, USA) containing a gelatin-coated polycarbonate membrane filter (8-μm pore size). The upper surface of the filter was coated with 20 μL Matrigel (0.3 mg/ml; BD Biosciences, Bedford, MA, USA). Transfected cells were harvested with a cell dissociation solution (Invitrogen, Carlsbad, CA, USA) and suspended in medium with 1% bovine serum albumin.

Cells (1 × 10⁴) were added to the upper compartment and allowed to migrate for 6 h at 37°C. After 6 h, cells on the upper side of the membrane were removed with a cotton swab. Cells on the bottom surface of the membrane were fixed in 3.7% paraformaldehyde, stained with hematoxylin for quantification. Data were expressed as the average number of cells per insert.

The expression of SLC38A1 protein in the healthy gastric mucosa was very low or undetectable (Figure 1A). Strong staining of SLC38A1 protein in the cytoplasm was found in 495 out of the 896 cancer samples in both the well differentiated (Figure 1B), and poorly differentiated cancer cells (Figure 1C). In the cytoplasm, the SLC38A1 protein was stained brown and diffused.

SLC38A1 overexpression was closely associated with old age (>60 yrs), differentiation status (high/moderate), regional lymph node metastasis, TNM stage (III), and positive PCNA expression (Table 1), but not with gender, tumor size, tumor location, operation manner, lymphovascular invasion, depth of tumor invasion, and p53 expression.

SLC38A1 expression in cancer tissue was associated with poorer prognosis. The median survival was 46.69 months in patients with SLC38A1-positive cancers, and 69.7 months in subjects with SLC38A1-negative cancers (P = 0.002; Figure 2). Upon a multivariate analysis, independent prognostic factors for overall survival included SLC38A1 expression, differentiation status, lymphovascular invasion, PCNA expression, size and TNM stage (Table 2).

Specific knockdown of endogenous SLC38A1 protein was confirmed by all four siRNAs with Western blot (Figure 3A). siRNA #2 (si-SLC38A1) reduced the endogenous SLC38A1 transcript by >70%, and was selected for subsequent experiments.

Compared to parental cells and si-Ctl, si-SLC38A1 decreased the proliferation (Figure 3B), as well as the migration of SH-10-TC cells (Figure 3C), but did not affect cell passage through Matrigel.

In the present study, we found an increased expression of SLC38A1 in gastric carcinomas, relative to adjacent non-cancerous gastric mucosa, suggesting that SLC38A1 might play an important part in gastric cancer malignant transformation. Enhanced SLC38A1 expression has been observed in several other types of malignancies, including liver cancer [6], Hilar cholangiocarcinoma [16] and C6 glioma [26]. An elevated expression of SLC38A1 has been found to be closely linked to differentiation status, lymph node metastasis and TNM staging, and is thus implicated in the growth, invasion, metastasis and progression of gastric carcinomas.

Consistent with the observation in HepG2 liver cancer cell lines [6], our study showed that inhibiting SLC38A1 expression with siRNA decreased the growth of cultured SH-10-TC cells, thus indicating an overexpression of SLC38A1 contributes to oncogenesis of gastric cancer through promoting cell proliferation. Inhibiting SLC38A1 expression also reduced cell migration, providing evidence for increased cell migration as a mechanism for enhanced metastatic potential, and local invasiveness of SLC38A1-expressing tumor cells.

In breast cancer, scientists discovered that 17 beta-estradiol specifically increased System A activity by two to four-fold in estrogen receptor positive cell lines, with a maximum stimulation observed 48 h after estrogen-treatment. In estrogen receptor negative cell lines, however, no stimulation was observed, which provided evidence that estrogen receptors play a role in the activation of system A by estrogen [27]. In liver cancer, System A amino acid transporters produced a different kind of inactivation and substrate protection in membrane vesicles and reconstituted proteoliposomes, supporting the hypothesis that there were inherent differences presented in System A carriers in normal and transformed liver tissue [28]. In the current study, we found enhanced SCL38A1 expression in gastric cancer, which supported the implication of glutamine metabolism in tumors [29‐31]. Interestingly, certain cancer cell lines are dependent on glutamine despite the fact that glutamine is a nonessential amino acid [32]. To be utilized by tumor cells, glutamine must be transported into tumor cell mitochondria. This implies that SLC38A1 could transport glutamine, and thus play an important role in neoplastic progression/development.

Upon further investigation of the relationship between SLC38A1 and clinicopathological factors, we found that overexpression of SLC38A1 was strongly associated with patient age, differentiation status, lymph node metastasis, TNM stage and the expression of PCNA, whereas no significant association was found in patient gender, tumor size, tumor location, operation manner, lymphovascular invasion, depth of tumor invasion, and p53 expression. These results indicated that SLC38A1 plays a central role in the malignant progression of gastric carcinomas.

Risk factors for gastric cancers have been reported in a number of studies, including differentiation, [33] nerve invasion [33], lymphovascular invasion [33], size [13, 34], TNM stage [34], and some molecular markers [35]. Our study confirmed that differentiation, lymphovascular invasion and the TNM stage were independent predictors for gastric cancer. We additionally found that SLC38A1 expression alone was a prognostic factor, with its prognostic value in multivariate survival analysis for patients who underwent gastrectomy. Therefore, we supposed that overexpression of SLC38A1 is a significant factor associated with a poor prognosis, and might be a marker to forecast gastric cancer patients’ recurrence and survival. Secondly, patients with poorer prognosis should be monitored more closely and/or undergo more positive treatment.