TEL2 suppresses metastasis by down-regulating SERPINE1 in nasopharyngeal carcinoma

Abstract

Metastasis is the major cause of treatment failure in patients with nasopharyngeal carcinoma (NPC). However, the molecular mechanisms of NPC metastasis are poorly understood. Here, using our customized gene microarray containing all of the known human transcription factors and the current markers for epithelial-mesenchymal transition, we report that TEL2 was down-regulated in highly metastatic NPC cells and the metastatic tissues in lymph node. Mechanistically, TEL2 inhibits the cell migration and invasion in vitro and metastasis in vivo by directly suppressing the SERPINE1 promoter in NPC. Consistently, an inverse correlation was observed between the protein levels of TEL2 and SERPINE1 using clinical NPC samples. Collectively, we have provided the first evidence that TEL2 plays a key role in NPC metastasis by directly down-regulating SERPINE1, and that this novel axis of TEL2 / SERPINE1 may be valuable to develop new strategies for treating NPC patients with metastasis.

Keywords: SERPINE1; TEL2; metastasis; nasopharyngeal carcinoma.

Figure 1. TEL2 is down-regulated in the metastatic cells and tissues of NPC

A. The expression profiles for human transcription factors and EMT markers in the indicted cell lines using the microarray as described in “Methods”. Primary: primary tumor tissues, LN: the metastatic tumor tissues in lymph node. B. The relative mRNA levels of TEL2 normalized to GAPDH level in the indicated cell lines, as determined by qRT-PCR. n = 3. The bars indicate the SD. *P < 0.05 and **P < 0.001 using Student's t-test. C. The proteins in the indicated cell lines were analyzed by Western blotting. D. The mRNA levels of TEL2 in the indicated tissues were measured as presented in b. P: NPC primary tissues, n = 10, L: the metastatic tumor tissues in lymph node, n = 4. The bars indicate the SD. Statistical analysis (Mann-Whitney test) was performed (P < 0.05). E, F. IHC for the clinical NPC samples of both primary sites and the metastatic tumor tissues in lymph node. The panel F presents representative images, while the panel E illustrates the statistical results using a Mann-Whitney test (P < 0.05). The dots represent the scores, while the bars indicate the SD. Primary NPC: primary NPC tissues, n = 138, LN metastases: the metastatic tumor tissues from NPC in lymph node, n = 16. Scale bars in F, 50 μm.

Figure 2. Overexpression of TEL2 suppresses migration and invasion of NPC cells

A. The proteins were analyzed in the indicted stable cell lines. B, C. Cell proliferation was measured in the indicted stable cell lines by the CCK8 assay. n = 3. D, E. Cell migration and invasion were determined in the indicated stable cell lines overexpressing TEL2 as described in “Methods”. n = 3. The bars indicate the SD. **P < 0.001 using Student's t-test. The representative images (100×) for the indicated stable cell lines were shown as the left panels.

Figure 3. knockdown of TEL2 promotes NPC metastasis

A. The proteins were analyzed in the indicted stable cell lines. B, C. Cell proliferation was measured in the indicted stable cell lines by the CCK8 assay. n = 3. D, E. Cell migration and invasion were determined in the indicated stable cell lines knockdown of TEL2 as described in “Methods”. n = 3. The bars indicate the SD. *P < 0.05 and **P < 0.001 using Student's t-test. The representative images (100×) for the indicated stable cell lines were shown as the left panels. F, G. The liver metastasis in vivo of the indicted stable cell lines in nude mice was determined as described in “Methods”. The panel F illustrates the statistical results. n = 6. The bars indicate the SD. **P < 0.01 and ***P < 0.0001 using Student's t-test. The panel G presents the six livers having the metastatic nodules from each group, scale bars, 1 cm.

Figure 4. Knockdown of TEL2 increased SERPINE1 level

A. Gene expression profiles using the stable cell line knockdown of TEL2 in S26 and 6–10B cells. B. The relative mRNA levels of SERPINE1 normalized to GAPDH level in the indicated cell lines, as determined by qRT-PCR (mean + SEM of triplicate samples are shown). C. The proteins in the indicated cell lines were analyzed by Western blotting. D. The relative mRNA levels of SERPINE1 in the indicated stable cell lines overexpressing TEL2 were determined as presented in B (mean + SEM of triplicate samples are shown). E. The relative mRNA levels of TEL and SERPINE1 in S18 cells transiently transfected with TEL or vector as determined (mean + SEM of triplicate samples are shown). F. The relative mRNA levels of SERPINE1 in the indicated cell lines were determined as presented in B (mean + SEM of triplicate samples are shown). G. The mRNA levels of SERPINE1 in the indicated tissues were measured as described in 1D. The bars indicate the SD. *P < 0.05 using Student's t-test.

Figure 5. TEL2 directly binds to and suppresses the SERPINE1 promoter activity

A. The luciferase activity of three regions of the SERPINE1 promoter in S26 cells was determined as specified in the Methods section. B. The schematic illustration of the suppression region within wild type SERPINE1 promoter and its mutants at both motif 1 and motif 2. C. The luciferase activity of wild type SERPINE1 promoter and its mutants in S26 cells after expressing TEL2 shRNA or scrambled shRNA. D. The indicated cells stably transfected with vector or HA-TEL2 as indicated were analyzed by ChIP assays using anti-HA as described in the Methods section.

Figure 6. The promotion of NPC metastasis induced by down-regulating TEL2 depends on the up-regulation of SERPINE1

A. Schematic illustration for the layout of TALEN target sites of SERPINE1. The left and right TALENs recognize the top and bottom strands of the target sites, respectively. B, C. the SERPINE1 protein and DNA deleted and disrupted in S26 cells was confirmed by Western blot (B) and DNA sequence (C), respectively. D. S26-SERPINE1 ^−/− cells were transfected with lentiviruses stably expressing scrambled shRNA or shRNAs targeting TEL2 at different locis (#1 and #3), and the cells were analyzed by western blotting. E. The cell migration and invasion of the S26-SERPINE1^−/− cells were determined as in 3(D) n = 3. The bars indicate SD. NS: non-statistical significance using Student's t-test. F. The liver metastasis in vivo of the S26-SERPINE1^−/− cells expressing TEL2 shRNA or scrambled shRNA in nude mice was determined as in 3(F, G). The six livers without metastatic nodules from each group are shown in the image. Scale bars, 1 cm. G, H. The cell migration and invasion were determined as in E. n = 3. The bars indicate the SE. *P < 0.05,**P < 0.001. NS: non-statistical significance using Student's t-test.

Figure 7. The clinical significance of the TEL2 / SERPINE1 axis in patients with NPC

A, B. IHC for TEL2 and SERPINE1 was performed using 138 clinical NPC samples as described in the Methods section, and an inverse correlation was observed between TEL2 and SERPINE1 in NPC samples in the clinic(P = 0.003, χ2 tests). Scale bars in B, 100 px C. The subgroup with TEL2 down-expression and SERPINE1 overexpression had lower overall survival compared to that with TEL2 overexpression and SERPINE1 down-expression. D, E. The subgroup with high SERPINE1 (D) or low TEL2 expression (E) had poor overall survival rate compared to those with low SERPINE1 (D) or high TEL2 expression (E). F. The concentration of SERPINE1 in serum of NPC patients with (35 samples) or without metastasis (35 samples) as indicated was measured by ELISA. The dots represent the scores, while the bars indicate the SD.