p53 targets TSPAN8 to prevent invasion in melanoma cells

Abstract

Cutaneous melanoma is a very deadly cancer because of its proclivity to metastasize. Despite the recent development of targeted and immune therapies, patient survival remains low. It is therefore crucial to enhance understanding of the molecular mechanisms underlying invasion. We previously identified tetraspanin 8 (TSPAN8) as an important modulator of melanoma invasiveness, and several of its transcriptional regulators, which affect TSPAN8 expression during melanoma progression toward an invasive stage. This study found that TSPAN8 promoter contains consensus-binding sites for p53 transcription factor. We demonstrated that p53 silencing was sufficient to turn on Tspan8 expression in non-invasive melanoma cells and that p53 acts as a direct transcriptional repressor of TSPAN8. We also showed that p53 modulated matrigel invasion in melanoma cells in a TSPAN8-dependent manner. In conclusion, this study reveals p53 as a negative regulator of Tspan8 expression. As TP53 gene is rarely mutated in melanoma, it was hitherto poorly studied but its role in apoptosis and growth suppression in melanoma is increasingly becoming clear. The study highlights the importance of p53 as a regulator of melanoma invasion and the concept that reactivating p53 could provide a strategy for modulating not only proliferative but also invasive capacity in melanoma treatment.

Figure 1

TSPAN8 functional promoter and p53 expression in melanoma cells. (a) Promoting sequence of the TSPAN8-002 isoform shows p53 consensus-binding sites as predicted by Genomatix GmbH, Munich, Germany. Promoter upstream of the transcriptional starting site (TSS) is colored in green, exons 1 and 2 in red and intron 1 in gray. The two putative half-sites for p53 binding are underlined, ‘site 1’ is colored in dark orange and ‘site 2’ in light orange. (b) Expression levels of TSPAN8 (left panel) and p53 (right panel) transcripts were assessed by reverse transcriptase (RT)–quantitative PCR (QPCR) in non-invasive IC8 and invasive T1C3 melanoma cells, previously described in Berthier-Vergnes et al. (n=3; ±s.e.m.). We checked that the TP53 gene was wild-type in these melanoma cell lines. The RT-QPCR protocol and primers used were described in Agaesse et al., except for p53 primers: p53-forward 5′-TGACTGTACCACCATCCACTA-3′ and p53-reverse 5′-AAACACGCACCTCAAAGC-3′. Statistical significance was calculated by a two-tailed Student's t-test for unpaired samples. Mean differences were considered significant when P<0.05 and ***P<0.001. NS, nonsignificant. (c) Expression levels of Tspan8 and p53 proteins were assessed by western blot in non-invasive IC8 and invasive T1C3 melanoma cells (n=3; a representative experiment is shown). Western blots were performed as previously described in. β-Actin (clone C4 Millipore 1/5000, Darmstadt, Germany) was used as a loading control. Tspan8 was detected using a mouse monoclonal anti-Tspan8 antibody (TS29.2 clone 1/2000) and p53 using a mouse monoclonal anti-p53 antibody (D01, Santa Cruz, Santa Cruz, CA, USA).

Figure 2

p53 silencing in non-invasive and invasive melanoma cells increases Tspan8 expression. (a, b) Quantitative PCR (QPCR) analysis showed p53 and TSPAN8 transcript expression levels 48 and 72 h after control or p53 small interfering RNA (siRNA) transfection in (a) non-invasive IC8 cells and (b) invasive T1C3 cells (n=3; ±s.d.). For siRNA transfection, 10⁵ cells per well were seeded in six-well plates and, after 24 h, transfected with 20 nM of control siRNA or p53 siRNA with INTERFERin (Polyplus, Illkirch, France). Targeting sequences were 5′-UAAGGCUAUGAAGAGAUAC-3′ for control siRNA and 5′-UAUGGCGGGAGGUAGACUG-3′ for p53 siRNA. (c) Expression levels of Tspan8 and p53 proteins were assessed by western blot in non-invasive IC8 (upper panel) and invasive T1C3 (lower panel) melanoma cells (n=3; a representative experiment is shown). Western blot quantifications were performed using ImageJ software (NIH/ImageJ, Bethesda, MD, USA). (d) Mean Tspan8 cell surface protein expression was assessed by fluorescence-activated cell sorting cytometry (FACS) in invasive T1C3 melanoma cells 72 h post-transfection with control or p53 siRNA, as previously described in Berthier-Vergnes et al. The left panel is representative of three independent experiments and the right panel represents the mean±s.d. of three independent experiments. (e) QPCR analysis showed p53 and TSPAN8 transcript expression levels 48 to 72 h after control or p53 siRNA transfection in non-invasive WM115 (left panel) and invasive SKMel28 (right panel) cells, in which TP53 gene is not mutated (n=3; ±s.d.). (f) The effect of nutlin-3 (5 μM; N-6287 Sigma-Aldrich, St Louis, MO, USA) on p53, p21 and Tspan8 expression in invasive T1C3 melanoma cells was assessed at 48 and 72 h post-treatment compared with control vehicle treatment (dimethylsulfoxide (DMSO)). Tspan8 mRNA levels (middle panel) were measured by QPCR (n=3; ±s.d.) and protein expression levels (upper and lower panel) of p53, p21 and Tspan8 were assessed by western blot (n=2; a representative experiment is shown). Statistical significance was assessed by two-tailed Student's t-test for unpaired samples. Mean differences were considered significant when P<0.05, *P<0.05 and **P<0.01.

Figure 3

p53 is recruited onto p53 consensus-binding sites in TSPAN8 promoter and represses Tspan8 expression. (a) p53-Chromatin immunoprecipitation (ChIP) assays were performed in invasive T1C3 melanoma cells (n=4). Enrichment of TSPAN8 promoter region (left panel) was analyzed in comparison with a negative control promoter region located −1-kb upstream of the beginning of TSPAN8 promoter. p21 Promoter region (right panel) was used as positive control. ChIP experiments were performed, as previously described by Masse et al. A representative experiment after agarose gel electrophoresis is shown in the lower panel. (b–e) Luciferase assays were performed with 20 nM of control or p53 small interfering RNA (siRNA) combined with 250 ng of (b) native pTSPAN8::LUC (constructed as described in Agaesse et al.) and/or (d, e) pTSPAN8 in which p53 consensus-binding sites were mutated. Directed mutagenesis was performed using the In Fusion HD Cloning Plus kit (Ozyme, Saint-Quentin-en-Yvelines, France), according to the manufacturer's instructions. PCR was performed at 55 °C. Primers used for mutagenesis were: forward 5′-TTCCGGGCCAAGTCCAGAGCATATTGCAGGA-3′ and reverse 5′-CTTGGCCCGGAACAGAGATTTCTGTATCCACG-3′ for ‘site 1’ forward 5′-CTTCGGGCAAGCTAACGAATAGTTAAATTCACGGC-3′ and reverse 5′-GCTTGCCCGAAGGCAATATGCTCTGGAGCA-3′ for ‘site 2’, and final sequences were described in c. For luciferase experiments, 10⁵ cells per well were seeded in 24-well plates and, after 24 h, transfected with 250 ng of control plasmid or plasmid of interest for 24 h, combined with 20 nM of control siRNA or p53 siRNA. Luciferase assays were performed using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) according to the manufacturer's instructions. The luminescence intensity ratio was calculated relative to that of the pGL4.10-empty-vector. Data were normalized to the transfection efficacy assessed by pCMV-RL vector co-transfection. At least three independent biological replicates were performed. Statistical significance was assessed by two-tailed Student's t-test for unpaired samples. Mean differences were considered significant when P<0.05, *P<0.05 and **P<0.01. NS, nonsignificant.

Figure 4

p53 modulates matrigel invasion in invasive cells in a Tspan8-dependent manner. (a–d) Boyden chamber matrigel invasion assays were performed on control or p53 small interfering RNA (siRNA) transfected invasive T1C3 (a) and SKMel28 (c) cells, on invasive T1C3 cells treated 48 h with 5 μM Nutlin-3 or dimethylsulfoxide (DMSO; control condition) (b) and on T1C3 cells stably expressing either control or TSPAN8 short hairpin RNA (shRNA) (d). In all, 10⁵ cells per well were seeded in six-well plates. Transfections or Nutlin-3 treatment were performed 24 h later with 20 nM of control or p53 siRNA or 5 μM Nutlin-3 or DMSO. After 36 h, cells were resuspended in 10% fetal calf serum (FCS)–McCoy’s medium by trypsination and centrifuged at 1200 g for 5 min. The residual FCS of the pellet was washed in 1 ml of phosphate-buffered saline and centrifuged. Finally, pelleted cells were resuspended in McCoy’s medium without FCS and for each condition, 35 000 cells were seeded in McCoy’s medium without FCS in the upper Boyden’s chamber, whereas the lower chamber was loaded with 5 μg/ml−10% FCS–McCoy’s medium. After 36 h of incubation, all the matrigel was scrubbed away, cells were fixated on the membrane for 15 min in −20 °C methanol, and stained for 10 min in 1 mM 4,6-diamidino-2-phenylindole (DAPI). The whole membrane surface was scanned at a magnification of × 100 using a time-lapse microscope scan slide protocol (ZEISS, Oberkochen, Germany), and all invaded cells were counted. Two (b, c; a representative experiment was shown) or three (a, d) independent biological replicates were performed. A representative visual field is illustrated in the right panel; scale bar=200 μM. Statistical significance was calculated by two-tailed Student's t-test for unpaired samples. Mean differences were considered significant when P<0.05, *P<0.05 and **P<0.01. NS, nonsignificant.