Selective participation of c-Jun with Fra-2/c-Fos promotes aggressive tumor phenotypes and poor prognosis in tongue cancer

Abstract

Tongue squamous cell carcinoma (TSCC) is most aggressive head and neck cancer often associated with HR-HPV infection. The role of AP-1 which is an essential regulator of HPV oncogene expression and tumorigenesis is not reported in tongue cancer. One hundred tongue tissue biopsies comprising precancer, cancer and adjacent controls including two tongue cancer cell lines were employed to study the role of HPV infection and AP-1 family proteins. An exclusive prevalence (28%) of HR-HPV type 16 was observed mainly in well differentiated tongue carcinomas (78.5%). A higher expression and DNA binding activity of AP-1 was observed in tongue tumors and cancer cell lines with c-Fos and Fra-2 as the major binding partners forming the functional AP-1 complex but c-Jun participated only in HPV negative and poorly differentiated carcinoma. Knocking down of Fra-2 responsible for aggressive tongue tumorigenesis led to significant reduction in c-Fos, c-Jun, MMP-9 and HPVE6/E7 expression but Fra-1 and p53 were upregulated. The binding and expression of c-Fos/Fra-2 increased as a function of severity of tongue lesions, yet selective participation of c-Jun appears to promote poor differentiation and aggressive tumorigenesis only in HPV negative cases while HPV infection leads to well differentiation and better prognosis preferably in nonsmokers.

Figure 1

(a–e): Constitutive activation and higher DNA binding activity of AP-1 in HPV^+ve and HPV^−ve tongue tumor tissues and cell lines. EMSA showing DNA binding activity of AP-1 in nuclear extracts (10 μg) of different grades of tongue tissues (a,b) and cell lines (c) using γ³²P- ATP-radiolabeled oligonucleotide harboring an AP-1 consensus sequence. Increasing AP-1 binding activity was observed as the severity of tongue lesions progressed from normal to precancer to invasive cancer in both HPV16^+ve (a) and HPV^−ve (b) cases. Both HPV16^+ve UPCI:SCC090 and HPV^−ve AW13516 (c) cell lines showed higher binding activity of AP-1. (d) EMSA with labeled Oct-1 probe showed uniform DNA binding in all grades of tongue tissues. Binding specificity was evidenced in a competition assay with nuclear extracts of TSCCs incubated with unlabelled 100 molar excess of specific competitor AP-1 probe and nonspecific competitor Oct-1 probe (e). Fold change in the band intensities of AP-1 are indicated in each lane. Ns: non-specific binding.

Figure 2

(a–d): Altered composition of functional AP-1 complex in HPV^+ve and HPV^−ve tongue tumor tissues and cell lines. Band supershift assay using specific antibodies (Abs) against all AP-1 family members (c-Jun, JunB, JunD, c-Fos, FosB, Fra-1, Fra-2). Panel a, b, c and d showing differential binding activity of AP-1 members in HPV^+ve (a) and HPV^−ve (b) TSCCs and tumor cell lines; UPCI:SCC090;HPV16^+ve(c) and AW13516; HPV^−ve(d). In all panels (a–d) significantly a higher (~90) binding of c-Fos and Fra-2 forms the functional AP-1 complex. HPV^+ve TSCC (panel a) shows a minor participation of JunD and JunB while c-Jun selectively participates in HPV^−ve tumors (panel b). The arrowhead indicates the supershifted bands. The intensities of super-shifted bands indicated and quantified in densitometric analysis. NS; non-specific binding.

Figure 3

(a–e): Differential expression pattern of AP-1 family proteins and their target genes in different grades of tongue tissues and cell lines. Representative immunoblots showing overexpression of AP-1 family proteins (c-Jun, JunB, JunD, c-Fos & Fra-2) and their target genes (Bcl-2, MMP-9 and Cyclin D1) in tongue tumor biopsies (a) and cell lines (b). 40 μg cellular protein from tongue tissues and cell lines were resolved on 10% SDS-PAGE, electrotransferred and probed with specific antibodies against all AP-1 family proteins and their target genes (Bcl-2, MMP-9 and Cyclin D1). To confirm equal loading of protein, the membranes were stripped and re-probed for β-actin expression and quantitation of bands was performed by densitometric analysis (lower panel of a & b). Note that there is significantly a higher expression of c-Fos & Fra-2 in tumor tissues, while Fra-1 is completely absent. Panel c; RT-PCR showing elevated mRNA expression profile of c-Jun, c-Fos & Fra-2 transcripts in tongue tumor tissues (c). GAPDH gene was used as internal loading control (lower panel of c). (Panel d,e); showing overexpression of Bcl-2, MMP-9 and Cyclin D1 genes in HPV^+ve and HPV^−ve tongue tumor tissues (d) and cell lines (e). β-actin gene was used as an internal control (lower panels of d,e). Significant (p < 0.05) when compared with control vs. cases and HPV^+ve vs. HPV^−ve cell lines. The data in bar diagram are expressed as the mean (±SEM or ±SD) of three independent experiments. *p < 0.05 & **p < 0.01, ***p < 0.001.

Figure 4

(a–e): Effect of Fra-2 silencing by specific siRNA in HPV^+ve and HPV^−ve TSCC cells. Representative immunoblots (Panel a,b) showing complete inhibition of Fra-2 expression at the concentration of 80 nM in both UPCI:SCC090;HPV16^+ve (a) and AW13516;HPV^−ve cells (b). 40 μg cellular proteins isolated from both the cell lines following treatment with Fra-2-siRNA (20 nM-80 nM) and scrambled control siRNA (80 nM) for 48 hours were examined for its effects on AP-1 family proteins (a,b), their target genes; Bcl-2, MMP-9 and CyclinD1 (c,d) and HPV16 E6/E7 (e). To ensure equal loading of protein, β-actin was used as an internal control. Quantitation of bands was performed by densitometric analysis. Significant (p < 0.05) when compared with untreated control vs. treated group. The data in bar diagram are expressed as the mean ± SD of three independent experiments. *p < 0.05 & **p < 0.01.

Figure 5

(a–d): Effect of Fra-2 silencing on cell invasion and migration of HPV^+ve and HPV^−ve TSCC cells. Fra-2 knockdown by specific siRNA leads to alterations in invasion in both HPV^−ve (a) and HPV^+ve (b) TSCC cells. Matrigel invasion assay of untreated, scrambled siRNA and Fra-2 siRNA treated in TSCC cells. Equal number of cells (1 × 10⁵) were seeded in the Matrigel coated upper chamber. After 24 hours of incubation at 37 °C and 5% CO₂, the cells that invaded through the membrane were fixed and stained with Giemsa, and images were captured using inverted microscope. Representative fields of view for each well are shown. Cell invasion was quantified by counting cells in six random fields. (c,d) Scratch assay of Fra-2 silencing demonstrated that cell migration was dramatically decreased in experimental groups after Fra-2 transfection compared with untreated and scrambled treated control groups. Inverted microscope images of wound closure at 0 hours, 12 hours and 18 hours in AW13516 (c) and UPCI:SCC90 (d) cells as indicated. Migration rate was calculated by image J software. The data in bar diagram are expressed as the mean ± SD of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 & ****p < 0.0001.