Cellular Entry of Human Papillomavirus Type 16 Involves Activation of the Phosphatidylinositol 3-Kinase/Akt/mTOR Pathway and Inhibition of Autophagy

Abstract

The mammalian target of rapamycin (mTOR) downstream of phosphatidylinositol 3-kinase (PI3K) in the growth factor receptor (GFR) pathway is a crucial metabolic sensor that integrates growth factor signals in cells. We recently showed that human papillomavirus (HPV) type 16 exposure activates signaling from GFRs in human keratinocytes. Thus, we predicted that the virus would induce the PI3K/mTOR pathway upon interaction with host cells. We detected activation of Akt and mTOR several minutes following exposure of human keratinocytes to HPV type 16 (HPV16) pseudovirions. Activated mTOR induced phosphorylation of the mTOR complex 1 substrates 4E-BP1 and S6K, which led to induction of the functional protein translational machinery. Blockade of epidermal GFR (EGFR) signaling revealed that each of these events is at least partially dependent upon EGFR activation. Importantly, activation of PI3K/Akt/mTOR signaling inhibited autophagy in the early stages of virus-host cell interaction. Biochemical and genetic approaches revealed critical roles for mTOR activation and autophagy suppression in HPV16 early infection events. In summary, the HPV-host cell interaction stimulates the PI3K/Akt/mTOR pathway and inhibits autophagy, and in combination these events benefit virus infection.

INTRODUCTION

Like many pathogens, human papillomavirus (HPV) entry into target cells is initiated by binding to cell surface heparan-sulfonated proteoglycans (HSPGs). The virus must then move to secondary receptors, which are responsible for particle internalization. Recently we showed that after interaction with HSPGs, HPV in complex with HS and growth factors (GFs) interacts with GF receptors (GFRs) and induces rapid activation of their pathways (1). Such receptors are often activated by viruses (2); the signals may be used to deceive the host's defenses, allowing safe entrance into the cell.

GFR activation triggers the phosphatidylinositol 3-kinase (PI3K)/Akt/mTOR signaling cascade, which is involved in controlling cellular macromolecular synthesis, metabolism, growth, and survival. Activated PI3K induces the conversion of phosphatidylinositol 4,5-bisphosphate (PIP₂) to phosphatidylinositol 3,4,5-trisphosphate (PIP₃), which recruits downstream factors to the cell membrane and regulates their activity (3). Akt is a key member of this pathway. PIP₃ anchors Akt to the plasma membrane, allowing its activation by phosphorylation. PIP₃ concentration is tightly controlled. The phosphatase PTEN negatively regulates PIP₃ concentration, converting PIP₃ to PIP₂ and thereby inhibiting PIP₃-mediated downstream signaling, including Akt activation. To prolong the infection cycle, viruses attempt to inhibit apoptosis and have developed several ways to activate Akt by enhancing the functions of the PI3K upstream regulator or by inhibiting negative regulatory phosphatases, or both (4, 5).

Akt's downstream effector, mTOR, is a crucial metabolic sensor, integrating diverse cellular signals that play critical roles in regulating many pathophysiological processes. This evolutionarily conserved serine/threonine protein kinase functions as a component of two structurally and functionally distinct signaling complexes: mTORC1 and mTORC2 (6). mTORC1 is activated by GFs and nutrients, regulates protein translation and cell growth, and plays an important role in the control of lipid synthesis (7) and mitochondrial metabolism (8). The best-characterized targets of mTORC1 are components of the translation machinery, including eukaryotic initiation factor (eIF)-4E-binding protein 1 (4E-BP1) and 40S ribosomal protein S6 kinase 1 (S6K1), both of which are important in the control of translation initiation (9). Sustaining the activity of mTORC1 is essential for continuing cap-dependent translation; therefore, viruses that rely on cap-dependent translation have acquired ways to prolong mTOR kinase activity (4). mTORC1 signaling can be potently inhibited by the naturally occurring antifungal macrolide rapamycin, which acts as an allosteric inhibitor (10) but does not completely inhibit mTORC1 activity (12). For this reason, PP242 and torin, recently discovered specific inhibitors capable of binding the catalytic site of mTORC1, are more widely used (14). The growth factor-sensitive but nutrient-insensitive mTORC2 phosphorylates Akt, SGK1, and PKC (12). These so-called “AGC group kinases” control multiple cellular functions, such as the structure of the actin cytoskeleton and cell survival (13–15). In contrast to mTORC1, mTORC2 is resistant to acute rapamycin treatment. Recent studies show that both mTORC1 and mTORC2 are involved in the regulation of autophagy (16, 17).

Autophagy is a tightly regulated cellular process responsible for eliminating damaged organelles, cell membranes, and proteins via a lysosomal pathway. Cell stress and diseases can also trigger this process. Cell autophagic machinery is known to capture and degrade intracellular pathogens (xenophagy in this case); this is an important component of the host response against viral infections (18). Therefore, many viruses have developed means to block autophagy or subvert this machinery (19).

HPVs regulate the PI3K/Akt/mTOR pathway, as do many other DNA viruses (5). HPV early proteins directly activate Akt and mTOR complexes (19–22). Here, we tested the activation of the PI3K/Akt/mTOR pathway upon keratinocyte exposure to HPV type 16 (HPV16) pseudovirions (PsVs), which deliver a reporter plasmid (pseudogenome) to the cell but express no viral proteins. HPV contact induces rapid activation of several signaling pathways in host cells (23–25), including that of PI3K/Akt potentially via alpha-6 beta-4 integrins (26). We recently demonstrated activation of epidermal GFR (EGFR) and keratinocyte GFR (KGFR) pathways in HaCaT cells exposed to HPV16 and 31 PsVs (1). Here we show that upon HPV16 PsV exposure to keratinocytes, EGFR signals result in rapid Akt activation by phosphorylation and by inactivation of PTEN. As a result, downstream events of this pathway were initiated: activation of mTOR and its substrates and suppression of autophagy. Biochemical and genetic approaches confirmed that these actions benefit virus infection.

MATERIALS AND METHODS

Cell culture, transfection, virus production, and infection quantification.

HaCaT and 293T cells were maintained as reported previously (1). HPV PsVs encapsidating a luciferase reporter plasmid were generated via transfection of 293T cells, CsCl gradient purified, and quantified for viral genome equivalents (vge) as previously described (27). Specifically, PsV preparations were isolated from a discrete and visible buoyant density band following CsCl gradient centrifugation. PsVs were desalted into HSB buffer (25 mM HEPES [pH 7.5], 0.5 M NaCl, 1 mM MgCl₂) using Amicon Ultra-4 100,000 MWCO centrifugation filter units (Millipore). PsV purity and L1 protein content were determined by SDS-PAGE and Coomassie brilliant blue staining against bovine serum albumin (BSA) standards (Fig. 1A). Virion morphology and quality were visualized by negative staining with 2% uranyl acetate and transmission electron microscopy (Hitachi 7500) at 80 kV following adsorption to a carbon-coated electron microscopy grid (Fig. 1B). Capsids reproducibly outnumber vge 2- to 10-fold (28). Infections were performed as reported earlier (see below) and allowed to proceed at 37°C, typically for 24 h before luciferase quantification (1). Luciferase quantities were normalized to protein content. For small interfering RNA (siRNA) depletions, HaCaT cells (20 to 25% confluence) were transfected with siRNA using Lipofectamine 2000 (Invitrogen). siRNAs targeting mTOR (Cell Signaling) or Beclin 1 or Atg7 (Dharmacon) were used according to the manufacturer's recommendations. A nonspecific siRNA was used as a negative control (Dharmacon). All infections were initiated using HPV16 PsVs at 100 to 200 vge/cell.

Fig 1.

Analysis of the purity of HPV16 PsV preparations. Following generation of HPV16 PsVs via transfection in 293T cells, the particles were subjected to buoyant density centrifugation in CsCl or Optiprep (iodixanol) gradients. The visible bands were collected by side puncture and desalted into HSB buffer using Amicon Ultra-4 100,000 MWCO centrifugation filter units (Millipore). (A) PsV purity and L1 protein content were determined by SDS-10% PAGE and Coomassie brilliant blue staining against bovine serum albumin (BSA) standards. The asterisk (*) may represent an L1 cleavage product. MW, molecular weight (in thousands). (B) Virion morphology and quality of PsVs purified by CsCl gradient visualized by negative staining and transmission electron microscopy. Bar, 100 nm.

Biochemical inhibition of infection and signaling.

For detection of phosphorylation, subconfluent HaCaT cells were serum starved for 4 h in Tyrode's buffer (10 mM HEPES [pH 7.4], 130 mM NaCl, 5 mM KCl, 1.4 mM CaCl₂, 1 mM MgCl₂, 5.6 mM glucose, and 0.05% BSA) and then were pretreated for 45 to 60 min with the indicated concentrations (see figure legends) of inhibitors in Tyrode's buffer: PP2 (Tocris Bioscience), PP242 (Cayman Chemical), LY294004 (Cell Signaling), AG1478 and PD168393 (Calbiochem), wortmannin, genistein, and rapamycin (Sigma). In experiments for detection of infection, inhibitors were dissolved in complete medium (Dulbecco's modified Eagle medium [DMEM] containing 10% fetal calf serum [FCS]) and incubated for 45 to 60 min with HaCaT cells grown in complete medium. HPV16 PsVs were added to inhibitor-pretreated cells, and infections proceeded in the presence of the inhibitors at 37°C, 24 h before luciferase quantification or for the indicated times prior to harvesting. Cell viability was assessed with each treatment by trypan blue exclusion staining.

SDS-PAGE and immunoblotting.

After addition of HPV16 PsVs, cells were incubated at 37°C for various times before being transferred to ice and solubilized with RIPA buffer (50 mM Tris [pH 7.5], 0.1% SDS, 1% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride [PMSF], 10 ng/ml leupeptin, 10 ng/ml aprotinin). Lysates were processed and analyzed by SDS-PAGE and proteins transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were probed with various monoclonal and polyclonal antibodies purchased from Cell Signaling: p-Akt (Ser-473), p-Akt (Thr-308), p-mTOR (Ser-2448), p-mTOR (Ser2481), mTOR, eIF4G, p-p70 S6K (Thr-389), p-p70S6K (Ser-371), p-PTEN (Ser-380), p-4E-BP1 (Thr-37/46), 4E-BP1, and LC3B. Other antibodies included PTEN (Abcam) and actin (MP Biomedical). Horseradish peroxidase-labeled secondary antibodies were used to detect antigen-antibody interactions (Pierce, Amersham Pharmacia Biotech). Immunoreactive bands were visualized by Super Signal West Pico Chemiluminescent Substrate (Thermo Scientific). For loading control analysis, blots were stripped and incubated with antibodies to nonphosphorylated proteins or actin. The bands were scanned and quantified using AlphaEaseFC (Alpha Innotech).

Immunofluorescence staining and confocal microscopy.

Immunofluorescence assays were performed essentially as described previously (1) using primary rabbit antibodies against p-PTEN (Ser-380), p-Akt (Thr-308), and p-Akt (Ser-473) (Cell Signaling) (1:150 dilution in 1% BSA–Tyrode) and secondary DyLight 488-conjugated Affinity Pure donkey anti-rabbit IgG (1:200; Jackson Immunochem.). For actin staining, phalloidin-rhodamine (1:500; Cytoskeleton, Inc.) was used. All images were acquired with a Zeiss LSM 510 META confocal system using appropriate filters. Parameters of laser intensities were kept constant during the imaging. Full-projection (three-dimensional [3D]) cell images were generated with Zen 2009 software (Zeiss), using z-stack confocal series. Fluorescent images in this paper were generated in the University of New Mexico (UNM) Cancer Center Fluorescence Microscopy Facility, supported as detailed on the webpage http://hsc.unm.edu/crtc/microscopy/index.html. Images were quantified with SlideBook 5.0.0.33 software and statistically analyzed using Prism software (GraphPad).

7-Methyl-GTP binding assays.

Subconfluent HaCaT cells were starved 4 h at 37°C in Tyrode's buffer prior to addition of HPV16 PsVs (200 vge/cell) for 1 h at 37°C. Cells were lysed (0.5% Triton X-100, 150 mM NaCl, 50 mM Tris [pH 7.5], 1 mM sodium orthovanadate; 1 mM PMSF; 10 ng/ml leupeptin). After centrifugation (10 min, 14,000 × g at 4°C), supernatants were incubated with a 50% slurry of 7-methyl-GTP-Sepharose (GE Healthcare, United Kingdom) prewashed in lysis buffer, and incubated 1 h at 4°C. After washing the resin two times with lysis buffer and twice with phosphate-buffered saline (PBS), samples were analyzed by SDS-PAGE and immunoblotting using anti-eIF4G rabbit polyclonal antibody (Cell Signaling). In experiments with inhibitors, cells were preincubated with 1 μM PD168393 (Calbiochem) or 5 μM PP242 (Cayman Chemical) before virus was added.

Cell cycle analyses.

HaCaT cells were incubated with compounds (inhibitors and siRNA) for 24 h. For detection of apoptotic cells, growing medium was harvested and floating cells were collected by centrifugation. The attached cells were washed with PBS and harvested using a trypsin-EDTA solution. Harvested cells were combined with apoptotic cells, washed 2 times with PBS, and fixed overnight in 70% ethanol. Cells were treated 30 min in the dark at room temperature with propidium iodide (PI) staining solution (0.1% TX100, 10 μg/ml PI, and 100 μg/ml RNase A in PBS). Samples were analyzed by flow cytometry at the UNM Center for Molecular Discovery.

RESULTS

HPV16 PsVs induce rapid activation of Akt in HaCaT cells.

Full Akt activation requires its translocation to the plasma membrane and dual phosphorylation of two distant segments of the polypeptide. Akt is phosphorylated at Thr-308 by PDPK1 and at Ser-473 by mTORC2 (13, 29). We tested whether HPV PsVs induced Akt activation upon interaction with HaCaT cells. Serum-starved cells were briefly exposed to HPV16 PsVs, and we detected rapid phosphorylation of Akt on both Thr-308 and Ser-473. Although mock-exposed controls had basal Akt phosphorylation at both sites (Fig. 2A, lane 1), exposure to HPV16 PsVs led to increased Akt phosphorylation, with levels reaching those achieved by EGF stimulation (Fig. 2A, lanes 2 and 6, respectively). The EGFR-specific inhibitor AG1478 markedly reduced the Akt phosphorylation levels, indicating that Thr-308 and, to a lesser extent, Ser-473 phosphorylation were a consequence of EGFR pathway activation (Fig. 2A, lane 3). Specific inhibitors of PI3K (LY294002 and wortmannin) abolished phosphorylation of both sites on Akt (Fig. 2A, lanes 4 and 5). To verify that Akt activation by PsVs was due to interaction with the cells, we employed heparin as an agent that prevents HPV binding to HaCaT cells (25, 30–32). When PsVs were saturated with heparin and unbound heparin was removed, Akt activation by HPV16 PsVs was blunted (data not shown). The well-characterized neutralizing antibodies H16.V5 and H16.U4 fail to avert HPV16 interaction with cells in culture (33) and also do not prevent Akt activation by HPV16 (data not shown). These findings provide further confirmation that HPV16 interaction with HaCaT cells triggers EGFR activation and also that HPV16 induces PI3K/Akt signaling very quickly upon cell contact. Confocal microscopy corroborated the rapid phosphorylation of Akt on both residues upon HPV16 exposure (Fig. 2B and C). The intensity of phospho-AKT on images was quantified, and t test analysis of mean intensity confirmed that the differences were significant (Fig. 2F).

Fig 2.

Activation of Akt and inhibition of PTEN by HPV16 PsVs. HaCaT cells were serum starved for 4 h. (A) Cells were treated with inhibitors (1 μM AG1478, 25 μM LY294002, or 1 μM wortmannin) or DMSO alone (Ø) for 1 h at 37°C. Cells were then exposed to 100 vge/cell HPV16 PsV for 15 min at 37°C in the presence of inhibitors. Cell lysates were fractionated by SDS-PAGE. Immunoblotting was performed using antibodies to p-Akt (Ser-473) and p-Akt (Thr-308). Serum-starved (4 h) HaCaT cells were incubated with 100 vge/cell HPV16 PsV, and the distribution of p-Akt (B and C) or PTEN (E) was assessed by confocal microscopy at 25 min postexposure. (D) Cells were mock treated or exposed to HPV16 for 25 min or 5 ng/ml EGF for 5 min. Postnuclear supernatants were analyzed by SDS-PAGE and immunoblotting for PTEN or p-PTEN (Ser-380). Numbers at the left of the gel represent kDa. Panel A includes lanes spliced together from the same exposure of the same films. Data are representative of 3 independent assays. (F) Akt and PTEN phosphorylation from panels B, C, and E was quantified with SlideBook; P values were determined by Student's t test.

HPV16 stimulates PTEN phosphorylation.

The PTEN phosphatase is a key inhibitor of the PI3K pathway, and PTEN phosphorylation on Ser-380 impedes its activity (34). When serum-starved cells were exposed to HPV16, microscopy revealed a strong increase of p-PTEN (Ser-380) compared to that seen in mock-exposed cells (Fig. 2E), and quantification of confocal microscopy images revealed a difference (Fig. 2F). Immunoblot analysis confirmed rapid Ser-380 phosphorylation of both 50-kDa and ∼100-kDa species of PTEN in response to HPV16 exposure. The higher-molecular-mass species specific to p-PTEN has not to our knowledge been described but was consistently detected as a dominant reactant even after samples were boiled 10 to 15 min under reducing conditions. We cannot exclude the possibility that the ∼100-kDa band is nonspecific. Nevertheless, levels of both the 50-kDa and ∼100-kDa p-PTEN forms following HPV16 exposure approached those induced by the Akt activator EGF (Fig. 2D, lane 3) (35). Densitometric analysis of the phospho-PTEN 50-kDa band in four independent experiments showed that phosphorylation doubled after HPV incubation with HaCaT cells. As Akt activation also could result from downregulated PTEN expression, we examined the total protein levels of PTEN in lysates from HPV-treated or untreated cells. The results showed a prominent PTEN species at 50 kDa and a minor form at ∼100 kDa (Fig. 2D); there was no significant difference in PTEN levels at either molecular mass in the presence of HPV16, indicating that the increase in Akt phosphorylation was not due to inhibited PTEN expression.

HPV16 rapidly activates mTOR.

Inactivation of PTEN by phosphorylation constitutively activates Akt, which then can phosphorylate and subsequently trigger mTORC1 (36, 37). mTOR activity involves phosphorylation on Ser-2448 via PI3K/Akt signaling and autophosphorylation of Ser-2481 (38, 39). As amino acids play a crucial role in mTOR activation, cells were starved in Tyrode's buffer to minimize nonspecific activation and then incubated with HPV16 for various times. mTOR was rapidly phosphorylated on both serine residues, and the phosphorylation levels were comparable to that observed upon cell stimulation by EGF or insulin (Fig. 3A, lanes 2 to 4 and lanes 5 and 6, respectively); Ser-2481 phosphorylation was not inhibited by rapamycin, even at high (μM) concentration (Fig. 3B, lanes 3 to 5). However, PP242 effectively blocked HPV16-induced mTOR phosphorylation on Ser-2481 residue at nM concentrations (Fig. 3B, lanes 6 to 8), suggesting that the mTORC1 complex is involved in HPV16-induced signal transduction.

Fig 3.

HPV16 PsVs induce phosphorylation of mTOR and activation of mTORC1 substrates S6K and 4E-BP1. HaCaT cells starved in Tyrode's buffer for 4 h were incubated with HPV16 PsVs. (A) Cells were lysed at the indicated times, and postnuclear supernatants were analyzed by SDS-PAGE and immunoblotting for total mTOR and phosphorylation-mediated activation of mTOR on Ser-2448 and Ser-2481. (B) Cells were treated for 1 h with DMSO (Ø, lane 2) or inhibitors in DMSO (0.5, 1, and 5 μM rapamycin in lanes 3, 4, and 5, respectively; 50, 500, and 5,000 nM PP242 in lanes 6, 7, and 8, respectively) prior to exposure to virus for 30 min. Immunoblot analysis was performed for total mTOR and p-mTOR (Ser-2481). Panels include lanes spliced together from the same exposure of the same films. (C to F) Cells were exposed to HPV16 PsVs for the indicated times before lysis and SDS-PAGE and immunoblot analysis of mTORC1 substrate S6K phosphorylation on Thr-389 (C) and pS6K Ser-371 (D) and substrate 4E-BP1 phosphorylation on Thr-37/46 (E and F); actin (C and D) or total 4E-BP1 (E and F) was detected as loading control. (F) HaCaT cells were pretreated for 1 h with DMSO (Ø), EGFR inhibitors (1 μM AG1478; 0.5 μM PD168393), PI3K inhibitors (25 μM LY294002; 1 μM wortmannin), mTOR inhibitors (3 μM PP242; 1 μM and 5 μM rapamycin), Src inhibitor (5 μM PP2), and a paninhibitor of tyrosine kinases (100 μM genistein) prior to virus exposure for 30 min. (G and H) HPV16 PsV-exposed cell lysates were subjected to a 7-methyl-GTP cap-binding pulldown assay as described in Materials and Methods and analyzed by SDS-PAGE and immunoblotting. (G) Actin was detected as a loading control. (H) Starved cells were incubated with DMSO (Ø), 5 μM PP242, or 1 μM PD168393 for 1 h before addition of HPV16 PsVs and incubation for an additional 1 h in the presence of inhibitors. Immunoblot detection of eIF4G shows equal amounts of this protein in lysates (used for pulldown assay). Panels A, C, E, and F include lanes spliced together from the same exposure of the same films.

HPV16 triggers mTORC1 substrates.

To more directly assess if HPV interaction with HaCaT cells increased mTORC1 activity, we evaluated the phosphorylation status of its two main substrates: S6K and 4E-BP1. Analysis of S6K revealed that HPV16 binding promotes rapid phosphorylation of both Thr-389 and Ser-371 residues, which are known to increase translational capacity by acting on translation initiation complex assembly (9, 40). Further, phosphorylation levels of these residues reached those observed after EGF stimulation of cells (Fig. 3C and D, respectively). Similarly, the level of phosphorylation of 4E-BP1 was markedly increased on Thr-37/Thr-46 residues and approached that induced by EGF (Fig. 3E). 4E-BP1 is phosphorylated on multiple sites, but phosphorylation of Thr-37/Thr-46 residues is a priming event catalyzed by mTOR (41, 42).

We examined whether phosphorylation of these translational regulators induced by HPV16 PsVs was connected to activation of the EGFR pathway. Inhibitors of EGFR (AG1478 and PD168393) and PI3K (LY294002 and wortmannin) completely prevented virus-induced 4E-BP1 phosphorylation, again demonstrating that HPV16 promotes phosphorylation of this protein as a result of EGFR pathway activation (Fig. 3F). Src kinases transduce signals from a variety of receptors (43, 44) and regulate PI3K/Akt activity (45). Thus, it was not surprising that inhibitors of Src kinases (PP2) and tyrosine kinases (genistein) also substantially prevented HPV16-mediated 4E-BP1 phosphorylation (Fig. 3F, lanes 8 and 11, respectively). To determine if the observed effect on 4E-BP1 phosphorylation was a result of mTORC1 activation, we treated cells with mTOR inhibitors rapamycin and PP242. PP242 substantially blocked phosphorylation of 4E-BP1 (Fig. 3F, lane 7). However, rapamycin, which is known to have little effect on the phosphorylation of 4E-BP1 (46, 47), inhibited phosphorylation of 4E-BP1 only at a very high concentration (Fig. 3F, lanes 9 and 10). These results again support the conclusion that HPV16 activates the PI3K/Akt/mTOR signal cascade, primarily via EGFR, in human keratinocytes.

HPV16 interaction with HaCaT cells leads to increased binding of eIF4G to a 7-methyl-GTP mRNA cap.

As phosphorylation of 4E-BP1 leads to the release of eIF4E to enable cap-dependent protein translation (48), we used an in vitro cap-binding assay to determine whether exposure to HPV16 PsV enhanced the assembly of the translation initiation complex at the mRNA cap. In three independent experiments, we observed increased eIF4G binding to the synthetic cap structure with lysates from HPV16 PsV-exposed cells compared to mock-exposed cells (Fig. 3G). To determine whether the increase of eIF4G binding to cap structures was caused by EGFR or mTORC1 activation, we performed similar pulldown experiments with cells treated with EGFR and mTOR inhibitors (PD168393 and PP242, respectively). As expected from the data above, both of these drugs inhibited eIF4G binding to the cap structure (Fig. 3H). Thus, it appears that the increased binding of translation initiation factors to the 7-methyl-cap is a result of EGFR/mTOR pathway activation catalyzed by HPV16 interactions with human keratinocytes.

HPV16 quickly suppresses autophagy.

In addition to activating biosynthetic processes, both mTORC1 and mTORC2 signaling paths are implicated in the inhibition of autophagy (49, 50). The serum and amino acid starvation conditions that we employed for mTOR phosphorylation detection induce autophagosome formation in HaCaT cells. Therefore, to determine whether HPV interaction with cells affects autophagy functions, we avoided starvation conditions and used HaCaT cells grown in complete medium (DMEM containing 10% FCS). HPV16 PsVs were incubated with cells for 1 h, and autophagy formation was evaluated by detecting the autophagosomal marker LC3. The levels of LC3-II, a form conjugated with phosphatidylethanolamine and present on membranes and autophagosomes, correlate with autophagy formation (51). As controls, cells were pretreated with 3-methyladenine (3MA), an inhibitor of autophagosome formation, or with tamoxifen, an autophagy inducer (52). Six independent experiments showed significantly decreased ratios of LC3-II to LC3-I (P value, 0.0145) in HPV16-exposed cell lysates compared to mock-exposed control cells (Fig. 4). This finding implies that HPV16 PsVs impair autophagosome formation early upon virus-host cell interaction.

Fig 4.

HPV16 interaction with keratinocytes suppresses autophagy. HaCaT cells grown in DMEM–10% FCS incubated with 100 vge/cell HPV16 PsVs and incubated at 37°C for 1 h. Controls included cells that were mock exposed to PsV, treated with 5 μM tamoxifen, or treated with 5 mM 3MA. (A) Cell lysates were analyzed by SDS-PAGE and immunoblotting for LC3, with actin detected as a loading control. Densities of LC3-I and LC3-II bands were analyzed with AlphaEaseFC software. (B) The graph represents data from six independent experiments quantifying the LC3-II/LC3-I ratio (two-tailed P value, 0.0145, indicating significance).

Activation of PI3K/Akt/mTOR signaling is critical for HPV16 infection.

The importance of the PI3K/Akt/mTOR pathway in HPV16 infection was assayed next. HaCaT cells were pretreated with specific inhibitors of this pathway and then infected with HPV16 PsVs in complete medium (Fig. 5A). The PI3K-specific inhibitor LY294002 reduced HPV16 pseudoinfection by ∼75%, which is in agreement with a previous report from our lab showing strong inhibition of HPV31 infection with wortmannin (53). Although wortmannin is 500-fold more potent than LY294002, it lacks specificity and in vitro can inhibit all three classes of the PI3K family, as well as mTOR, DNA-PK, some phosphatidylinositol 4-kinases, myosin light-chain kinase, and mitogen-activated protein kinase (MAPK) at high concentrations (54). Thus, the LY294002 data more specifically demonstrate the involvement of PI3K in infection. As predicted from the data in Fig. 3B, rapamycin had a negligible effect on HPV16 infection even at 1 μM, whereas the selective mTOR inhibitor PP242 significantly reduced virus infection with a 50% effective concentration (EC₅₀) of 2 ×10⁻⁶ M (Fig. 5B). The Src family kinase inhibitor PP2 also had a negative effect on HPV16 PsV infection (Fig. 5A), as anticipated from the fact that Src regulates PI3K/Akt and signal transduction from a variety of receptors (43–45). PP2 caused a dose-dependent reduction of infectivity with an EC₅₀ of 1.5 × 10⁻⁶ M (Fig. 5C). To verify that PP2 and PP242 did not inhibit HPV infection by impeding HPV attachment, HaCaT cells were incubated with dimethyl sulfoxide (DMSO) or inhibitors for 1 h at 4°C. Unbound virus was washed out, lysates were fractionated by SDS-PAGE, and the amount of HPV16 bound to the cells was determined by immunoblotting. Figure 5D shows that PP2 and PP242 did not alter the capacity of HPV to bind cells, indicating that the compounds do not inhibit HPV16 infection by impairing viral attachment to the cell surface. Similarly, the tyrosine kinase inhibitor genistein had no effect on HPV binding to cells. Again, these findings are consistent with the activation of GFRs by HPV16.

Fig 5.

Chemical inhibitors have varied effects on HPV16 infectivity. HaCaT cells were pretreated for 1 h with DMSO (Ø) or inhibitors at the indicated concentrations. Cells were exposed to HPV16 PsVs or mock exposed (M), and infectivity was scored after 24 h of incubation at 37°C. (A) Infection in the presence of 5 μM PP2, 50 μM LY294002, 3 μM PP242, 100 μM rapamycin, 5 mM 3MA, or 10 μM tamoxifen. (B and C) Dose-response curves of infection to mTOR inhibitor PP242 (B) or Src kinase inhibitor PP2 (C). The 50% infective concentrations (IC₅₀s) were calculated using GraphPad Prism software. (D) Immunoblot analysis of HPV16 L1 bound to HaCaT cells in the presence or absence of 5 μM PP2, 100 μM genistein, or 3 μM PP242. (E and F) Kinetics of HaCaT cell infection inhibition by PP242 and PP2. Cells were treated with DMSO (Ø), 4 μM PP242 (E), or 5 μM PP2 (F) for the indicated times pre- and postattachment of HPV16 PsVs (100 vge/cell), which was defined as 0 h; infection was analyzed 24 h following HPV exposure.

PI3K and Src activation are required at early steps of HPV infection.

We next strove to identify the temporal window in which PI3K and Src signaling is important for infection and thus the times that PP242 and PP2, respectively, inhibit HPV infection. HaCaT cells were incubated in the presence of DMSO, PP2, or PP242 for 1 h prior to HPV16 PsV addition or for various intervals after virus exposure. Cell infectivity was scored after 24 h. Addition of PP242 or PP2 at 4 to 5 μM concentrations prior to virus exposure inhibited infection 85% and 72%, respectively (Fig. 5E and F, respectively). Supplying these compounds to cells at time intervals greater than or equal to 1 h after incubation with virus was increasingly less inhibitory, demonstrating that PI3K and Src signals are important very early in infection.

mTOR functions and autophagy are important for HPV infection.

A genetic approach was used to interfere with mTOR and autophagosome functions to avoid possible off-target effects caused by biochemical inhibitors. First, HaCaT cells were transfected with either mTOR-specific or negative-control siRNAs. mTOR knockdown in three separate experiments ranged from ∼20 to 50% compared to cells transfected with a nonspecific control siRNA (Fig. 6A). Remarkably, HPV16 PsV infection levels in three independent experiments were reduced 80 to 90% compared to negative-control-siRNA-transfected cells (Fig. 6B). As additional confirmation of mTOR knockdown and inhibited function, phosphorylation levels of 4E-BP1 were only ∼15% of those levels observed in nonspecific-siRNA-transfected cells (Fig. 6C). These data show that mTOR functions are needed for successful HPV16 infection of human keratinocytes. Inhibition of mTOR complexes is known to activate autophagy (55). Analysis of LC3-II/LC3-I levels confirmed the initiation of autophagy in mTOR-depleted HaCaT cells (Fig. 6D and E). Consequently, the observed inhibition of infection upon mTOR depletion could result from viral degradation by autophagy. To test this hypothesis, we activated or blocked autophagy in HaCaT cells using tamoxifen or 3MA, respectively. Induction of autophagy with tamoxifen completely abrogated HPV16 infection, whereas infection was enhanced up to 6-fold when autophagy was inhibited with 3MA (Fig. 5A). As these widely used autophagy regulators are not exclusive to the autophagy machinery (56–61), we depleted key proteins of autophagosome formation, Beclin 1 and Atg7, using siRNAs. Immunoblot analysis revealed the successful knockdown of these proteins (Fig. 6F). Infection assays showed significant activation of infection compared to cells treated with nonspecific siRNA (Fig. 6G). As mTOR signaling affects translation, we verified that inhibitor-induced reductions in luciferase levels in the pseudoinfection assays were not due to suppressed translational events. Assaying the levels of reporter gene RNAs by reverse transcription- and quantitative PCR demonstrated that the luciferase transcript levels were consistent with luciferase protein levels (data not shown), suggesting that the inhibitors block infection prior to nuclear delivery and transcription of pseudogenomes. Thus, the genetic approach using siRNAs confirmed the results obtained from the biochemical analysis: autophagy inhibits HPV16 early infection, and interfering with autophagy results in enhanced HPV16 infection of host keratinocytes. Together with the data shown in Fig. 4, these results suggest that autophagy plays an antiviral role in HPV infection and that HPVs have means of interfering with this cellular process to promote infection.

Fig 6.

Autophagy plays a negative role in HPV infection. HaCaT cells were transfected with a control siRNA or siRNAs specific to mTOR, Beclin 1, or Atg7. Downregulation of these proteins was determined in transfected cells by immunoblot analyses using antibodies against mTOR (A) and Beclin 1 and Atg7 (F). Immunoblotting was used to determine the effect of mTOR depletion on p-4E-BP1 (C) and autophagy formation detected by analysis of LC3 (D). The LC3-II/LC3-I ratio (E) was quantified from three independent experiments by densitometry, and P values were determined by Student's t test. Actin (A, D, F) or 4E-BP1 (C) was detected as loading control. Infections were initiated 24 h posttransfection and proceeded for 24 h thereafter under normal conditions (B, G). Panel A includes lanes from the same exposure of the same films.

Cell cycle progression and mTOR inhibition.

Because progression into early M phase is needed for HPV infection (62), we tested whether the inhibitors prevented infection via cell cycle blockade. Therefore, we assayed the fraction of cells in each phase of the cell cycle during the inhibitor treatments under which infection levels were determined above. Although each inhibitor affected the cell cycle distribution, in no case did an inhibitor arrest the cells in any one phase of the cell cycle. No correlation was seen between infection inhibition and cell cycle distribution under the assay conditions employed (Fig. 7A). For example, the distributions of cells in the G₁, S, or G₂/M phases of the cell cycle were relatively similar whether cells were grown in complete medium and infected with HPV16 with no treatment or treated with 4 μM PP242 or 5 μM tamoxifen. However, infection levels ranged from 0% inhibition with no inhibitor to nearly 99% inhibition with tamoxifen (Fig. 5A). Specifically, the moderate changes observed in the number of cells in G₂/M phase were not sufficient to account for the levels of infection inhibition demonstrated for a given inhibitor. To determine whether mTOR inhibitors were preventing infection by arresting the cell cycle in G₁ or S phase, PP242-treated cells were incubated with monastrol, which increases the number of cells in late M phase and promotes HPV16 infection (1, 62). Cells were also pretreated with monastrol and then incubated with PP242 before exposure with HPV16 PsV. Treatment of cells with monastrol in all cases increased the number of cells in G₂/M phase (Fig. 7A). Monastrol alone significantly increased infection of cells with HPV16, and as expected we observed a 2-fold increase in infectivity (Fig. 7B). PP242 addition in the presence of monastrol did not alter the monastrol-characterized cell cycle profile, but infection was dramatically inhibited, by ∼85% (Fig. 7). These data indicate that cell cycle effects cannot account for the inhibition of infection by PP242 compounds. We expected to obtain an increase of G₁ phase after depletion of mTOR by transfection of HaCaT cells with mTOR-specific siRNA, but interestingly two independent experiments showed an increase of S phase. It should be noted that the number of apoptotic cells after mTOR depletion increased and reached ∼15% of total cell number (Fig. 7A).

Fig 7.

Cell cycle distribution fails to correlate with HPV16 infection inhibition. (A) HaCaT cells were transfected with mTOR siRNA or treated with drugs (50 μM LY294004, 1 μM PP242, 100 μM monastrol, 1 μM rapamycin, 10 μM tamoxifen, or 5 mM 3MA). Some cells were first incubated for 6 h with 1 μM PP242 and then with 100 μM monastrol (PP242+Monast); other cells were first incubated for 6 h with 100 μM monastrol and then with 1 μM PP242 (Monast+PP242). After 24 h of incubation at 37°C, cells were assayed with propidium iodide and examined using flow cytometry. The fractions of cells in G₁ (1n), S (intermediate), and G₂/M (2n) phases and apoptotic cells were expressed as percentages of the total cells counted. (B) HPV16 PsV infection levels (24 h postinfection) in the presence of complete media (CM) or inhibitors following pretreatment for 1 h with 100 μM monastrol, pretreatment with monastrol for 1 h plus 3 μM PP242 for the duration (Monast+PP242), or pretreatment with 3 μM PP242 for 1 h plus 100 μM monastrol for the duration (PP242+Monast).

DISCUSSION

We recently reported that HPV16 PsVs induce rapid activation of GF-regulated pathways, particularly EGFR and KGFR (1). EGFR activation by HPV16 PsVs was independently confirmed (25). EGFR signaling in most mammalian cells initiates activation of the PI3K/AKT/mTOR pathway; therefore, we examined if HPV triggers activation of these pathways downstream of EGFR. We found that HPV16 interaction with human keratinocytes initiates signal cascades in the PI3K/Akt/mTOR pathway to facilitate cap-dependent translation, inhibit autophagy, and thus bolster the early infection process.

Two important aspects of virus-induced PI3K/Akt/mTOR activation were revealed herein. First, HPV16 initiates PI3K/Akt/mTOR activity prior to viral early protein expression, perhaps before virion entry. Many viruses pose as routine activating ligands to induce normal cell signaling, avoid antiviral activity, and initiate safe uptake (2). However, activation of the PI3K/Akt pathway is typically associated with de novo viral protein expression aimed to hinder apoptosis and prolong infection (63). Most, if not all, mammalian DNA viruses modulate PI3K activity by changing the ratio of PIP₂ to PIP₃. Some viruses encode proteins capable of activating PI3K, and others induce cellular proteins having similar functions. Repression of PTEN activity or its expression is another means whereby viruses modulate PIP₃ concentration (4). HPV early proteins are well-known facilitators of PI3K-Akt signaling. HPV16 E5, E6, and E7 inhibit apoptosis by activating Akt (20–22). Interestingly, even after early HPV gene expression ensues, EGFR signaling plays a vital role in infection. For example, in human keratinocytes, HPV16 immortalization leads to heightened EGFR activity (64) and expression of HPV16 E5 induces dimerization of EGFR and subsequent upregulation of PI3K-Akt signaling (19).

We show that HPV interaction with host cells induced rapid phosphorylation of Akt and PTEN, leading to the activation of Akt by a two-pronged approach: phosphorylation of Akt kinase as well as phosphorylation of PTEN and, consequently, inactivation of this negative regulatory phosphatase. In addition, Akt activation is sensitive to EGFR-specific inhibition, indicating that HPV16 interaction with EGFR initiates this pathway. Our results are in good agreement with a publication from Fothergill and McMillan showing rapid activation of PI3K/Akt within several minutes of HPV or bovine PV virus-like particles (VLPs) binding to a serum-starved A431 epidermoid carcinoma cell line (26). Phosphorylation of Akt on Ser-473 residue was markedly increased upon virus binding, while only minor phosphorylation of Akt on Thr-308 was observed (26). Phosphorylation of Akt on Ser-473 is an indicator of mTORC2 complex activation. Our data show that HPV interaction with host cells changes the functionality not only of mTORC2 but also of another mTOR-regulated complex, mTORC1. Fothergill and McMillan attributed PI3K/Akt signaling to VLP interaction with alpha-6 beta-4 integrin (26). Interestingly, there is signaling cross talk on epithelial cells among EGFR, alpha-6 beta-4 integrin, and tetraspanin CD151 (65–67), each of which has been shown to be important for mediating HPV early infection (1, 68, 69). Thus, these molecules may be cooperating during HPV entry into keratinocytes.

In the very early stages of HPV16 interaction with host cells, activation of mTORC1 was evidenced by phosphorylation of mTORC1 and its substrates, 4E-BP1 and S6K, both of which are important in the physiological control of translation initiation. A primary function of mTORC1 is regulating cap-dependent mRNA translation (70, 71). Thus, it is not surprising that multiple DNA viruses, HPV among them, have evolved mechanisms to keep mTOR complexes active during infection. Phosphorylation of S6K increases the translational capacity by acting on translation initiation complex assembly (9, 40), and phosphorylation of 4E-BP1 induces the dissociation of this protein from eIF4E, allowing eIF4G to bind and form the eIF4F complex, which enhances cap-dependent initiation of mRNA translation (9). We observed each of these phosphorylation events after HPV16 interaction with host cells under conditions of nutrient deprivation. The pulldown assay with 7-methyl GTP cap analogue detected increased cap-dependent translation after only 1 h of virus binding to host cells, indicating changes in translation even before the bulk of virus internalization. Specific inhibitors of EGFR (PD168393) and mTOR (PP242) strongly inhibited eIF4G binding to 7-methyl GTP Sepharose 4B beads, demonstrating inhibition of translation and the crucial role of the EGFR/mTOR pathway in this process.

The second remarkable finding of this work is that signaling downstream of HPV16-GFR contacts early after virus-host interactions serves to inhibit rather than activate autophagy to promote more-efficient infection. HPV16 PsV exposure to HaCaT cells, by virtue of activation of mTOR complexes, opposes the formation of autophagosomes. This is significant, as autophagy is an important regulator of the host response against viral infections, wherein cells capture and degrade intracellular pathogens (18). Many viruses have developed a variety of ways to block autophagy or subvert this machinery for their own benefit (19). However, of those viruses that inhibit autophagy, all do so by de novo expression of virulence factors that impair the function of autophagy proteins. For example, the DNA viruses of the Herpesviridae family encode diverse proteins that inhibit autophagy by varied means (73, 76–80).

Cell engagement by several viruses has been found to activate autophagy (19, 72–74); however, our findings show that the initial HPV-cell interaction leads to autophagy inhibition to promote infection. Furthermore, HPV16 infectivity was directly affected by modulating keratinocyte autophagy. Infection levels were particularly enhanced upon inhibiting autophagy with 3MA. This finding is in good agreement with work from the Pyeon laboratory, showing an ∼40-fold increase of HPV16 infectivity in primary keratinocytes in the presence of 3MA (81). Tamoxifen, an inducer of autophagy, practically abolished infection. Yet both tamoxifen and 3MA impact targets other than those related to autophagy (56–61). The specificity of genetic knockdowns of mTOR, Beclin 1, and Atg7 confirmed that HPV-induced restraint of autophagy is important for early infection events.

In summary, we report here that HPV16 PsV interaction with HaCaT human keratinocytes causes rapid activation of PI3K/Akt/mTOR signaling and the impairment of autophagy and that both of these processes benefit HPV16 infection. This signaling cascade was primarily influenced by activation via the EGFR. Recently, we showed stimulation of the MAPK (ERK1/2) pathway after HPV16 interaction with EGFR and KGFR on HaCaT cells (1). Cross talk between ras/MAPK and PI3K/Akt effector paths generates context-dependent responses to growth factor stimulation, alters the dynamic topologies of signal propagation networks downstream of cell surface receptors, and leads to amplification or attenuation of key target protein activities (75). Thus, HPV interaction with cell surface GFRs is likely to trigger analogous communication between these pathways, resulting in activation of complementary signaling, cellular evasion of apoptosis, enhanced viral transcription and translation, promoted proliferation, and prolonged infection. As many other viruses are capable of interacting with EGFR at the cell surface, we propose that these viruses might also induce the PI3K/Akt/mTOR pathway via this GFR activation to inhibit autophagy upon viral uptake.