E6AP is essential for the proliferation of HPV-positive cancer cells by preventing senescence

Abstract

Oncogenic types of human papillomaviruses (HPVs) are major human carcinogens. The formation of a trimeric complex between the HPV E6 oncoprotein, the cellular ubiquitin ligase E6AP and the p53 tumor suppressor protein leads to proteolytic p53 degradation and plays a central role for HPV-induced cell transformation. We here uncover that E6AP silencing in HPV-positive cancer cells ultimately leads to efficient induction of cellular senescence, revealing that E6AP acts as a potent anti-senescent factor in these cells. Thus, although the downregulation of either E6 or E6AP expression also acts partially pro-apoptotic, HPV-positive cancer cells surviving E6 repression proliferate further, whereas they become irreversibly growth-arrested upon E6AP repression. We moreover show that the senescence induction following E6AP downregulation is mechanistically highly dependent on induction of the p53/p21 axis, other than the known pro-senescent response of HPV-positive cancer cells following combined downregulation of the viral E6 and E7 oncoproteins. Of further note, repression of E6AP allows senescence induction in the presence of the anti-senescent HPV E7 protein. Yet, despite these mechanistic differences, the pathways underlying the pro-senescent effects of E6AP or E6/E7 repression ultimately converge by being both dependent on the cellular pocket proteins pRb and p130. Taken together, our results uncover a hitherto unrecognized and potent anti-senescent function of the E6AP protein in HPV-positive cancer cells, which is essential for their sustained proliferation. Our results further indicate that interfering with E6AP expression or function could result in therapeutically desired effects in HPV-positive cancer cells by efficiently inducing an irreversible growth arrest. Since the critical role of the E6/E6AP/p53 complex for viral transformation is conserved between different oncogenic HPV types, this approach could provide a therapeutic strategy, which is not HPV type-specific.

Fig 1. Downregulation of E6AP induces senescence in HPV-positive cancer cells.

(A) Treatment scheme: HeLa or SiHa cells were reverse transfected (Tx) with three different E6AP-specific siRNAs (siE6AP-1, -2, -3), or control siRNA (siNeg). 72 h post-transfection, cells were either harvested for protein (Prot) or RNA analyses, or split and further cultivated for senescence assays (SA-β-Gal staining) or colony formation assays (CFAs) after the indicated time periods. (B) Immunoblot analyses of E6AP, E6, E7, p53, p21, and Vinculin protein levels. (C) qRT-PCR analyses of E6AP mRNA levels. Shown are log2-transformed fold changes (log2FC) of mean expression with standard deviations (n = 3). Statistically significant differences between cells transfected with siE6AP-1, -2, or -3, and those transfected with siNeg (log2FC = 0) were assessed using one-way ANOVA with Sidak’s test for multiple comparisons. *** p ≤ 0.001. (D) Corresponding senescence assays (upper panels; SA-β-Gal staining, blue; scale bar: 200 µm) and CFAs (lower panels).

Fig 2. Downregulation of E6AP or E6 exerts pro-apoptotic effects in HPV-positive cancer cells.

HeLa or SiHa cells were transfected with 20 nM of siE6AP, si18E6 or si16E6, or control siRNA (siNeg), and cultivated under either 1 g/L or 4.5 g/L glucose in the cell culture medium for 96 h. Upper panels: TUNEL analyses of HeLa or SiHa cells (scale bar: 50 µm). Lower panels: Quantification of TUNEL positive (apoptotic) cells relative to the number of DAPI-stained cells. Results are presented as mean percentages with standard deviations (n = 15 fields of view, each with ≥ 50 cells from three independent experiments). Statistically significant differences for cells transfected with siNeg compared to siE6AP or siE6 and for cells cultivated under 1 g/L or 4.5 g/L glucose were assessed using two-way ANOVA with Tukey’s test for multiple comparisons. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Fig 3. Differential effects of downregulating E6AP, E6/E7, or E6 on the senescence response of HPV-positive cancer cells.

HeLa or SiHa cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg) and further examined following the treatment scheme depicted in Fig 1A. Shown are corresponding senescence assays (upper panels; SA-β-Gal staining, blue; scale bar: 200 µm) and CFAs (lower panels).

Fig 4. Differential effects of downregulating E6AP, E6/E7, or E6 on the cell cycle regulation of HPV-positive cancer cells.

(A) HeLa-mKate2 or SiHa-mKate2 cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg) and imaged by the Incucyte S3 live-cell imaging system for up to 144 h post-transfection. Cell counts (red objects) were quantified using the Incucyte software package and are shown relative to the counts at 24 h post-transfection (set to 1). Rel, relative; Tx, transfection. (B) HeLa or SiHa cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg), cultivated for 72 h, and treated with nocodazole or solvent control (-) for the last 16 h (HeLa) or 24 h (SiHa) of cultivation. Cell cycle profiles and quantifications of the percentages of cell populations in the individual cell cycle phases are shown. (C) HeLa or SiHa cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg), cultivated for 72 h, and examined by immunoblot for B-MYB, FOXM1, E2F1, Cyclin A, Cyclin B1, CDC2, CDK2, CKS1, E6AP, E6, E7, and GAPDH protein levels. (D) Corresponding qRT-PCR analyses of MYBL2 (coding for B-MYB), FOXM1, CCNA2 (coding for Cyclin A2), and E2F1 mRNA levels. Shown are log2-transformed fold changes (log2FC) of mean expression with standard deviations (n = 3). Statistically significant differences between cells transfected with siE6AP, siE6/E7, or siE6, and those transfected with siNeg (log2FC = 0) were assessed using one-way ANOVA with Sidak’s test for multiple comparisons. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Fig 5. p53 and p21 are critical for senescence induction upon E6AP downregulation.

(A) HeLa or SiHa cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg), cultivated for 72 h, and examined by immunoblot for p53, P-p53 Ser15, P-p53 Ser20, Ac-p53 Lys382, p21, E6AP, E6, E7, and GAPDH protein levels. (B) qRT-PCR analyses of CDKN1A (coding for p21) mRNA levels. Shown are log2-transformed fold changes (log2FC) of mean expression with standard deviations (n = 3). Statistically significant differences between cells transfected with siE6AP, siE6/E7, or siE6, and those transfected with siNeg (log2FC = 0) were assessed using one-way ANOVA with Sidak’s test for multiple comparisons. ** p ≤ 0.01, *** p ≤ 0.001. (C-E) HeLa or SiHa cells were transfected with siE6AP, si18E6/E7 or si16E6/E7, si18E6 or si16E6, or control siRNA (siNeg), alone or in combination with sip53 or sip21, as indicated. Cells were examined following the treatment scheme depicted in Fig 1A by (C) immunoblot for p53, p21, E6AP, E6, E7, and GAPDH protein levels (D) senescence assays (SA-β-Gal staining, blue; scale bar: 200 µm), and (E) corresponding CFAs.

Fig 6. p53 and p21 are critical for cell cycle changes upon E6AP downregulation.

(A) HeLa-mKate2 or SiHa-mKate2 cells were transfected with siE6AP or control siRNA (siNeg), either alone or in combination with sip53 or sip21, as indicated, and imaged by the Incucyte S3 live-cell imaging system for up to 120 h post-transfection. Cell counts (red objects) were quantified using the Incucyte software package and are shown relative to the counts at 24 h post-transfection (set to 1). Rel, relative; Tx, transfection. (B) Cell cycle analyses of HeLa or SiHa cells, transfected with siE6AP or control siRNA (siNeg), either alone or in combination with sip53 or sip21, as indicated. Cells were cultivated for 72 h and treated with nocodazole or solvent control (-) for the last 16 h (HeLa) or 24 h (SiHa) of cultivation. Cell cycle profiles and quantifications of the percentages of cell populations in the individual cell cycle phases are shown.

Fig 7. Downregulation of E6AP does not induce senescence in “p53 null” cells.

HeLa or HeLa “p53 null” cells (HeLa A1, HeLa C2) were transfected with siE6AP, si18E6/E7, or control siRNA (siNeg). Cells were examined following the treatment scheme depicted in Fig 1A by (A) immunoblot for p53, p21, E6AP, 18E6, 18E7, and GAPDH protein levels (B) senescence assays (SA-β-Gal staining, blue; scale bar: 200 µm), and (C) corresponding CFAs.

Fig 8. Senescence induction upon E6AP or E6/E7 downregulation depends on p130 and pRb.

(A) SiHa cells were transfected with siE6AP, si16E6/E7, si16E6, or control siRNA (siNeg). After 72 h, protein levels of P-pRb Ser807/811, total pRb, p130, E6AP, 16E6, 16E7, and β-Actin were measured by immunoblot analyses. (B,C) SiHa cells were transfected with siE6AP, si16E6/E7, or control siRNA (siNeg), either alone or in combination with sip130, sipRb, or a combination of both sip130/sipRb, as indicated. Cells were examined following the treatment scheme depicted in Fig 1A by (B) immunoblot for P-pRb Ser807/811, total pRb, p130, E6AP, 16E6, 16E7, and β-Actin protein levels and by (C) senescence assays (SA-β-Gal staining, blue; scale bar: 200 µm). (D) SiHa cells were transfected with siE6AP, si16E6/E7, si16E6, or control siRNA (siNeg), either alone or in combination with sip53 or sip21. Protein levels of P-pRb Ser807/811, total pRb, and GAPDH were assessed 72 h post-transfection by immunoblot analyses.

Fig 9. Tentative model for the pro-senescent effect of E6AP downregulation.

In HPV-positive cancer cells, E6AP forms a trimeric complex with E6 and p53, leading to the proteolytic degradation of p53 (upper panel). Formation of this complex is abolished upon downregulation of E6AP, and E6 is destabilized (lower panel). This leads to a strong upregulation of p53 and p21 levels, and subsequent inhibition of the phosphorylation of p130 and pRb through p21-mediated suppression of CDKs. Both p130 and pRb are important for senescence induction upon E6AP downregulation, e.g., by being a key component of the multiprotein DREAM complex (p130) or by inhibiting E2F (pRb). Notably, senescence induction upon E6AP downregulation can take place in the presence of E7 (please refer to the discussion in the text). Created with BioRender.com.