High-risk human papillomavirus E6 oncoproteins interact with 14-3-3ζ in a PDZ binding motif-dependent manner

Abstract

Cervical cancer develops through the combined activities of the human papillomavirus (HPV) E6 and E7 oncoproteins. A defining characteristic of E6 oncoproteins derived from cancer-causing HPV types is the presence of a PDZ binding motif (PBM) at the extreme carboxy terminus of the protein which is absent from E6 proteins derived from the so-called low-risk HPV types. Within this PBM is also a protein kinase A (PKA) phospho-acceptor site, which is thought to negatively regulate the association of E6 with its PDZ domain-containing substrates. We can now show that phosphorylation of E6 by PKA and/or AKT confers the ability to interact with 14-3-3ζ. The interaction is direct and specific for the high-risk HPV E6 oncoproteins, although there are significant differences in the efficiencies with which HPV-16, HPV-18, and HPV-31 E6 oncoproteins can associate with 14-3-3ζ; this correlates directly with their respective susceptibilities to phosphorylation by PKA and/or AKT. We demonstrate here that the interaction between E6 and 14-3-3ζ also requires integrity of the E6 PBM, and downregulation of 14-3-3ζ results in a marked reduction in the levels of HPV-18 E6 expression in HeLa cells. Using phospho-specific anti-E6 antibodies, we also demonstrate significant levels of E6 phosphorylation in vivo. These studies redefine the potential relevance of the E6 PBM in the development of cervical cancer, suggesting that interaction with 14-3-3ζ, as well as the more well-established interactions with PDZ domain-containing substrates, is likely to be responsible for the biological activities attributed to this region of the high-risk HPV E6 oncoproteins.

Fig 1

Amino acid sequence of HPV-18 E6 and its consensus recognition sites. The wild-type HPV-18 E6 sequence around residues S82 and T156 is shown. Upper section, consensus recognition sequences for phosphorylation with PKA (47) and AKT (49). Also shown is the consensus recognition motif for 14-3-3 (30) and the PDZ binding motif (48). Lower section, E6 mutants used in the study. Potential phospho-acceptor residues are shown in red, phospho-mimic mutations are shown in green, and Ala substitutions are shown in blue.

Fig 2

High-risk HPV E6 oncoproteins are phosphorylated to differing degrees by PKA. (A) The purified GST fusion proteins were either untreated or incubated with PKA and [γ-³²P]ATP (circled P) as indicated. Proteins were then subjected to SDS-PAGE and autoradiographic analysis. Upper panel, autoradiogram from a series of parallel reactions; lower panel, the Coomassie blue-stained gel. (B and C) PKA phosphorylation of different HPV-18 E6 mutant-GST fusion proteins. Upper panels, autoradiograms; lower panels, the Coomassie blue-stained gels. Arrows indicate the relevant GST proteins. The residual levels of phosphorylation for the T156D and R153A mutants are approximately 1% of the level obtained with the wild-type E6 fusion protein.

Fig 3

Comparative analysis of E6 phosphorylation by different kinases. (A through C) Phosphorylation of GST fusion proteins of HPV-18 E6, the HPV-18 E6 T156E mutant, and Scribble (Scrib) with PKA, CamKII, AKT, and PAK, as indicated. Upper panels, autoradiograms; lower panels, Coomassie blue-stained gels. (D and E) Phosphorylation of GST fusion proteins of HPV-18, HPV-16, and HPV-31 E6s and the HPV-18 E6 R153A mutant by AKT. Upper panels, autoradiograms; lower panels, the Coomassie blue-stained gels. Arrows indicate the relevant GST fusion proteins.

Fig 4

Detection of phospho-E6 in vivo. (A) The indicated GST fusion proteins were either untreated or subjected to phosphorylation with PKA (+) in the presence of nonradiolabeled ATP. Proteins were then detected by Western blotting with anti-E6 phospho-specific antibody. The lower panel, shows Ponceau staining of the nitrocellulose membrane, confirming equal levels of protein loading in each assay. (B and C) HEK293 cells were transfected with empty vector or the indicated HA-tagged E6 expression plasmids in the presence (+) or absence (−) of forskolin (Fsk). After 24 h, the cells were harvested and subjected to immunoprecipitation using anti-HA-conjugated agarose beads. The presence of E6 was then detected by Western blotting using either the anti-E6 phospho-specific antibody to detect phosphorylated E6 (upper panels) or the anti-HA antibody to detect total levels of E6 protein (lower panels).

Fig 5

HPV E6 interacts with 14-3-3ζ in a phosphorylation-dependent manner. (A) The indicated GST fusion proteins were either untreated or subjected to phosphorylation with PKA (circled P) in the presence of nonradiolabeled ATP. They were then incubated with in vitro-translated radiolabeled 14-3-3ζ, MAGI-1, or Dlg as indicated. Following extensive washing, bound proteins were detected by using SDS-PAGE and autoradiography. Upper panel, autoradiogram; lower panel, the Coomassie blue-stained gel. The arrows indicate the relevant fusion proteins and translated products. (B) The indicated GST fusion proteins were either untreated or subjected to phosphorylation with PKA (circled P) in the presence of nonradiolabeled ATP. They were then incubated with in vitro-translated radiolabeled 14-3-3ζ. Following extensive washing, the bound 14-3-3ζ was detected using SDS-PAGE and autoradiography. Upper panel, autoradiogram; lower panel, the Coomassie blue-stained gel. Arrows indicate the relevant proteins. (C) Assays to monitor 14-3-3ζ interactions with the different HPV-18 E6 mutant-GST fusion proteins were performed as described for panel B.

Fig 6

Direct interaction between HPV E6 and 14-3-3ζ. (A) Interaction assay with purified 14-3-3ζ. Purified GST fusion proteins were either untreated or subjected to phosphorylation with PKA (circled P) in the presence of nonradiolabeled ATP. They were then incubated with purified 14-3-3ζ. After extensive washing, the bound protein was detected by Western blotting using anti-His antibody (upper panel). Ponceau staining of the nitrocellulose membrane was also performed (lower panel). (B) Results of quantitation from at least three independent assays. (C) As described for panel A, except phosphorylation was performed with AKT. (D) Results of quantitation from at least three independent assays.

Fig 7

14-3-3ζ contributes to maintaining HPV-18 E6 steady-state levels. (A) HeLa cells were transfected with control siRNA (siLuc) or siRNA against 14-3-3ζ. After 72 h, the cells were harvested, and the levels of E6 were ascertained by Western blotting using anti-HPV-18 E6 monoclonal antibody. Also shown are the levels of 14-3-3ζ and the loading control α-actinin. (B) Results of quantitation from at least three independent experiments showing the levels of 14-3-3ζ knockdown and the effects on E6 levels.