RASSF1A inhibits PDGFB-driven malignant phenotypes of nasopharyngeal carcinoma cells in a YAP1-dependent manner

Abstract

Nasopharyngeal carcinoma (NPC) is a highly aggressive tumor characterized by distant metastasis. Deletion or down-regulation of the tumor suppressor protein ras-association domain family protein1 isoform A (RASSF1A) has been confirmed to be a key event in NPC progression; however, little is known about the effects or underlying mechanism of RASSF1A on the malignant phenotype. In the present study, we observed that RASSF1A expression inhibited the malignant phenotypes of NPC cells. Stable silencing of RASSF1A in NPC cell lines induced self-renewal properties and tumorigenicity in vivo/in vitro and the acquisition of an invasive phenotype in vitro. Mechanistically, RASSF1A inactivated Yes-associated Protein 1 (YAP1), a transcriptional coactivator, through actin remodeling, which further contributed to Platelet Derived Growth Factor Subunit B (PDGFB) transcription inhibition. Treatment with ectopic PDGFB partially increased the malignancy of NPC cells with transient knockdown of YAP1. Collectively, these findings suggest that RASSF1A inhibits malignant phenotypes by repressing PDGFB expression in a YAP1-dependent manner. PDGFB may serve as a potential interest of therapeutic regulators in patients with metastatic NPC.

Fig. 1. Expression of RASSF1A correlates with the self-renewal properties and tumorigenicity of NPC cells.

a Protein expression levels of RASSF1A in NPC cells; β-actin was used as the loading control. b–d CNE-2 cells stably transfected with overexpressing RASSF1A (RF1) or with empty vector (Vec) were analyzedas follows. (b, left panel) RASSF1A protein expression levels were determined by western blot analysis; β-actin was used as the loading control. c A cell proliferation curve was constructed from MTS assay results, the data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test. d Single-cell suspensions were seeded in ultra-low-attachment culture plates. The formed spheroids were counted via microscopy, and representative images are shown. The representative images and numbers of RASSF1A-overexpressing and control cells were compared, *p < 0.05, Student’s t test. Scale bar: 200 µm. b, e, f CNE-1 cells stably transfected with shRNA targeting RASSF1A (sh2, sh5) or scrambled shRNA (shLuc) were analyzedas follows. (b, right panel) Western blot analysis of RASSF1A expression; β-actin was used as the loading control. e A cell proliferation curve was constructed from MTS assay results, the data are presented as the mean ± S.D. values, *p < 0.05 and **p < 0.01 for CNE-1/sh2 cells compared with CNE-1/shLuc cells; ^##p < 0.01 for CNE-1/sh5 cells compared with CNE-1/shLuc cells; Student’s t test. f Single-cell suspensions were seeded in ultra-low-attachment culture plates. The formed spheroids were counted via microscopy, and representative images of CNE-1 cells transfected with RASSF1A shRNAs or negative control scrambled shRNA are shown. g A total of 2 × 10⁵ (upper) and 5 × 10⁴ (lower) RASSF1A-overexpressing and its control CNE-2 cells were subcutaneously injected into NOD/SCID mice (n = 10 mice/group). Summary of tumorigenicity in mice were shown. The TIF and p value were calculated using ELDA software. h A total of 2 × 10⁵ (upper) and 5 × 10⁴ (lower) RASSF1A shRNA- or scrambled shRNA-targeting CNE-1 cells were subcutaneously injected into NOD/SCID mice (n = 10 mice/group). Summary of tumorigenicity in mice were shown. The TIF and p value were calculated using ELDA software.

Fig. 2. RASSF1A regulated expression of molecules involved in cell movement and ultimately inhibits the motility and invasive phenotype of NPC cells.

a Western blot analysis of E-cadherin and vimentin in RASSF1A-overexpressing (left panel) CNE-2 cells, RASSF1A-depleted (right panel) CNE-1 cells and their corresponding control cells, GAPDH was used as the loading control. b RASSF1A markedly attenuated the migration and invasion characteristics of CNE-2 cells with or without presence of mitomycin C(1 ug/ml). Representative images of the migration assay (upper) and invasion assay (lower) of RASSF1A-overexpressing and control CNE-2 cells. The data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test. Scale bar: 100 µm. c RASSF1A depletion enhanced CNE-1 cell migration and invasion with or without presence of mitomycin C (1 ug/ml), as determined by migration (upper pannels) and invasion (lower pannels) assays, the data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test. Scale bar: 100 µm.

Fig. 3. PDGFB is crucial for maintaining malignant properties induced by RASSF1A in NPC cells.

a A heat map generated using the significantly changed genes categorized in the “cytokine-cytokine receptor interaction pathway” is shown. b, c mRNA expression (b) was evaluated by qRT-PCR and protein concentration by ELISA (c) in CM of RASSF1A-overexpressing CNE-2 cells, RASSF1A-depleted CNE-1 cells and their corresponding control cells, The data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test. d–g PDGFB was transiently knocked down with a pool of siRNA or treated with a neutralizing antibody for PDGF-BB (10 µg/mL) in RASSF1A-depleted CNE-1 cells. PDGF-BB secretion in the CM was measured by ELISA (d), **p < 0.01, Student’s t test. e Number of spheroids formed was determined via microscopy, and representative images (e left panel) are shown. The formed spheroids were compared (e right panel), the data are presented as the mean ± S.D. values, *p < 0.05, **p < 0.01, Student’s t test; ns: non-sinificant. Scale bar: 200 µm. Representative images of the migration assay (f) and invasion assay (g) are shown, the data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test. Scale bar: 100 µm. h–j Recombinant PDGF-BB or IgG was added to RASSF1A-overexpressing CNE-2 cells. Representative images of sphere formation (h) (Scale bar: 200 µm.), migration (i) and invasion (j) assays (Scale bar: 100 µm) of RASSF1A-overexpressing CNE-2 cells treated with PDGF-BB (the culture medium was supplemented with 20 ng/ml or an equal volume of control IgG) are shown, The data are presented as the mean ± S.D. values, *p < 0.05, **p < 0.01, Student’s t test.

Fig. 4. RASSF1A induces actin cytoskeletal rearrangement and inhibits YAP1 activation.

a Representative images of F-actin stained with phalloidin and observed by immunofluorescence microscopy. Nuclei were visualized by DAPI staining, scale bar: 25 µm. b YAP1 expression in the total, cytosolic or nuclear fractions of RASSF1A-overexpressing CNE-2 cells, RASSF1A-delepted CNE-1 cells and their corresponding control cells were determined by western blot analysis. c mRNA quantification of CYR61 and CTGF by using qRT-PCR; GAPDH was used as the internal control. The data are presented as the mean ± S.D. values, *p < 0.05, **p < 0.01, Student’s t test.

Fig. 5. PDGFB induction by YAP1 mediates the modulatory effect of RASSF1A on the malignant phenotypes of NPC cells.

a–c YAP1 was transiently knocked down in RASSF1A-depleted CNE-1 cells. a In the indicated cells, YAP1 protein expression was assessed by using western blotting; b PDGFB, CYP61 and CTGF mRNA expression was assessed by qRT-PCR; c Concentration of PDGF-BB secreted in CM was measured by ELISA. The data are presented as the mean ± S.D. values, *p < 0.05, **p < 0.01, Student’s t test. d–f Recombinant PDGF-BB or IgG was added to YAP1-silenced NPC cells. d The formed spheroids were counted via microscopy, and (e) representative images are shown. f The impact of PDGF-BB treatment on the migration and invasion of RASSF1A-depleted cells was determined by Transwell assays. The data are presented as the mean ± S.D. values, *p < 0.05, **p < 0.01, Student’s t test. Scale bar: 200 µm.

Fig. 6. RASSF1A inactivates YAP1 through actin rearrangement.

a–c RASSF1A-depleted CNE-1 cells were treated with or without LTB (10 µM) treatment for 1 h, and its control CNE-1 cells were also included into assays as follows. a Representative images of F-actin stained with phalloidin were observed by immunofluorescence microscopy. Nuclei were visualized by DAPI staining, scale bar: 25 µm. b YAP1 expression in the total, cytosolic or nuclear fractions was determined by western blot analysis. c mRNA quantification of YAP1, CYR61, CTGF and PDGFB by using qRT-PCR; GAPDH was used as the internal control. The data are presented as the mean ± S.D. values, **p < 0.01, Student’s t test.