Fig 2: TIPRL attenuates migration and invasion of gastric cancer cells. (A–D) The MKN45 and BGC823 cells were transfected with TIPRL-expressing vector [pcDNA3.1 (+)-TIPRL], TIPRL siRNA, and the respective control vector. The protein levels of TIPRL were examined by WB (***P < 0.001). (E,F,I,K) Expression of TIPRL inhibited the migration and invasion of MKN45 and BGC823 cells, as determined by transwell migration and matrigel invasion assays (t-test, *P < 0.05; **P < 0.01). (G,H,J,L) siRNA-mediated knockdown of TIPRL promoted the migration and invasion of MKN45 and BGC823 cells (t-test, *P < 0.05; **P < 0.01). All experiments were performed in triplicate. Error bars, SD.

Fig 3: Identification and expression evaluation of TIPRL as a metastasis-related gene in gastric cancer. (A,B) Microarray plots of metastatic tumors (M group) vs. non-metastatic tumors (N group) demonstrating a different expression profile. Hierarchical cluster analysis shows 372 genes are significantly altered in gastric cancer samples with or without lymphatic metastasis (A, P < 0.05 with >2-fold change, raw value >500). Four novel genes (TIPRL, SERINC3, COPS6, and SELM) are significantly differentially expressed in metastatic gastric cancer tissues compared with their counterparts (B). Each column represents a sample; each row denotes the expression level of a single gene. Expression level is demonstrated by colors: Green, underexpressed genes; red, overexpressed genes. (C–F) The mRNA expression levels of TIPRL, but not SERINC3, COPS6, and SELM, were significantly lower in gastric cancer tissues with LNM (positive) than those without LNM (negative), as determined by RT-qPCR (t-test, P = 0.0209, P = 0.3707, P = 0.1220, and P = 0.0689, respectively). (G–I) Low mRNA expression of TIPRL was correlated with poor overall survival (G, P = 1.1e-5), first progression survival (H, P = 8.5e-5), and post-progression survival (I, P = 1.7e-5) in gastric cancer patients from Kaplan-Meier plotter (http://kmplot.com/).

Fig 4: Decrease of TIPRL expression correlates with a poor clinical outcome in gastric cancer. (A–C) Representative immunohistochemical staining for TIPRL expression in normal gastric tissues, gastric cancer with different TNM tumor stages (I-II vs. III-IV) and gastric tumors with or without distant metastasis (M0 vs. M1). Original magnification, × 100. (D) Statistical analysis of TIPRL expression in gastric cancer tissues. (E) Quantitative analysis of TIPRL staining indicated that staining intensity in gastric tumors was significantly lower than normal gastric mucosa (t-test, ***P < 0.001). (F) TIPRL expression was dramatically decreased in patients at advanced stages (III-IV), in contrast with those at earlier stages (I-II) (t-test, *P < 0.05). (G) Analysis of TIPRL staining intensity also showed lower staining intensity in patient samples with distant metastasis (M1) compared to those without distant metastasis (M0) (t-test, *P < 0.05). (H) The ROC curves demonstrated strong separation between normal and gastric cancer tissues [AUC = 0.6490, CI (95%): 0.5709–0.7270, P = 0.0004]. The sensitivity and specificity of TIPRL expression to distinguish normal from gastric cancer tissues were 43.27 and 82.56%, respectively. (I,J) Kaplan–Meier survival curves revealed that low intensity of TIPRL immunostaining strongly correlated with poor overall survival (I, log-rank test, P = 0.0193) and disease-free survival (J, log-rank test, P = 0.0364).

Fig 5: Effect of TIPRL on cell apoptosis in gastric cancer cells. (A–H) Up-regulation or down-regulation of TIPRL expression could not induce apoptosis in both cell lines, as indicated by flow cytometry analysis following Annexin V-FITC/PI staining (t-test, P > 0.05). All experiments were performed in triplicate. Error bars, SD.