Nuclear Entry of High-Risk Human Papillomavirus Type 16 E6 Oncoprotein Occurs via Several Pathways

Preparation of recombinant human nuclear import factors.

His-tagged Kap α2/hSRP1α (58) and His-tagged Kap β1/p97 (4) were expressed in Escherichia coli BL21(DE3) (3-h induction with 2 mM isopropyl-β-d-thiogalactopyranoside [IPTG] at 30°C), and the soluble His-tagged proteins were purified in their native state on Talon beads using a standard procedure. GST-Kap β1 (5) and GST-Kap β2 (6) were expressed in E. coli BL21(DE3) (3-h induction with 1 mM IPTG at 30°C), and the soluble GST fusion proteins were purified in their native state on glutathione-Sepharose beads using a standard procedure. To obtain cleaved Kap β2, the GST-Kap β2 fusion protein was incubated for 2 h at room temperature with a Tev enzyme that has a His tag (Invitrogen), and after cleavage the GST was removed by binding to glutathione-Sepharose beads and the Tev enzyme was removed by binding to Talon beads, using standard procedures. Human Ran (7) was prepared as described previously (12). All the proteins were checked for purity and lack of proteolytic degradation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie blue staining. The purified proteins were dialyzed in transport buffer (20 mM HEPES-KOH [pH 7.3], 110 mM potassium acetate, 2 mM magnesium acetate, 1 mM EGTA, 2 mM dithiothreitol, plus protease inhibitors) and stored in aliquots at −80°C until use.

Preparation of recombinant HPV16 E6 oncoprotein.

The recombinant full-length HPV16 E6 oncoprotein (20, 21) was expressed as a GST fusion protein in E. coli BL21(DE3) and purified in native conditions on glutathione-Sepharose beads using a standard procedure.

Preparation of GST-NLS fusion proteins and GST-E6 mutants.

The C-terminal NLS of the HPV16 E6 (₁₂₁KKQRFHNIRGRWTGRCMSCCRSSRTRRETQL₁₅₁) and the E6 R124G NLS mutant were fused to a GST reporter in the following manner. The forward and reverse oligonucleotides consisted of the following base pairs. For GST-HPV16 E6 _121-151: forward, 5′-/5Phos/GGA ATT CCT GGA CAA AAA GCA AAG ATT CCA TAA TAT AAG GGG TCG GTG GAC CGG-3′; reverse, 5′-/5Phos/GCG CGC TCG AGT TAC AGC TGG GTT TCT CTA CGT GTT CTT GAT GAT CTG CAA CAA GAC ATA CAT CGA CCG GTC CAC CGA CCC C-3′). For the GST-HPV16 E6_121-151 mutant (R124 to G): forward, 5′-/5Phos/GGA ATT CCT GGA CAA AAA GCA AGG ATT CCA TAA TAT AAG GGG TCG GTG GAC CGG-3′; reverse, 5′-/5Phos/GCG CGC TCG AGT TAC AGC TGG GTT TCT CTA CGT GTT CTT GAT GAT CTG CAA CAA GAC ATA CAT CGA CCG GTC CAC CGA CCC C-3′. The oligonucleotides were designed with a 15- to 20-bp overlapping GC-rich area and annealed equimolarly. The overhanging areas were filled in using DNA polymerase I, large Klenow fragment (New England Biolabs catalog no. M0210S). The filled-in Klenow fragments were blunt end ligated with T4 DNA ligase. These fragments were double digested with EcoRI and XhoI and inserted into pGEX 4T-1 (Amersham Pharmacia Biotech 27-4580-01) at a vector-to-insert ratio of at least 1:10. To prepare the E6_R124G mutant and the E6_ΔKKQR deletion mutant, the QuikChange site-directed mutagenesis kit from Stratagene was used according to the manual's instructions. All the constructs were transformed into XL1 blue bacteria and confirmed by automated sequence analysis (BioServe Biotechnologies Sequencing Department). GST-mpNLS_HPV16L1 was prepared as previously described (35), and the GST-M9 construct was a gift from Gideon Dreyfuss. For protein expression, the constructs were used to transform E. coli BL21(DE3). After induction of transformed E. coli BL21(DE3) with 1 mM IPTG for 3 h at 37°C, the GST-NLS fusion proteins were purified in their native state on glutathione-Sepharose beads using a standard procedure. The purified proteins were checked by SDS-PAGE and Coomassie blue staining, dialyzed in transport buffer, and stored in aliquots at −80°C until use.

In vitro nuclear import assays.

We carried out in vitro nuclear import assays in digitonin-permeabilized HeLa cells, as previously described for HPV L1 major capsid proteins (34, 35). HeLa cells are cervical carcinoma cells that contain HPV18 DNA integrated into the cellular genome with consequent disruption of several viral genes but preservation of the E6 and E7 genes. HeLa cells contain low levels of HPV18 E6, localized by immunofluorescence with specific antibodies (sc-1585 from Santa Cruz Inc., Santa Cruz, Calif.) both in the nucleus and in the cytoplasm (data not shown). Although HeLa cells are HPV18 positive, this does not affect the import properties of their nuclear pore complexes, as demonstrated by the fact that many mammalian nuclear import pathways have been identified and characterized with HeLa cells (1, 4, 15, 25, 37, 39, 41, 42, 48, 58). Briefly, subconfluent HeLa cells, grown on poly-l-lysine (Sigma P-4707)-coated glass coverslips for 1 day, were permeabilized with 70 μg of digitonin/ml for 5 min on ice. Unless otherwise specified, the import reactions contained an energy regenerating system (0.5 mM GTP, 5 mM phosphocreatine, and 0.4 U of creatine phosphokinase) plus various transport factors (1 μM Kap α2; 0.5 μM Kap β1; 0.5 μM Kap β2; 3 μM RanGDP), plus the GST-E6 or the GST-NLS fusion protein (0.5 μM). The final import reaction volume was adjusted to 20 μl with transport buffer. At the end of the nuclear import reaction, the cells were fixed with 3.7% formaldehyde on ice followed by methanol for 3 min at −20°C. In order to prevent nonspecific binding, cells were blocked for 1 h in 3% bovine serum albumin with 0.1% Tween in phosphate-buffered saline. For visualization of nuclear import, the GST fusion proteins were detected by immunofluorescence with an anti-GST antibody (1:200 dilution; Amersham Pharmacia 27-4577-01) followed by a fluorescently labeled secondary antibody (1:100 dilution; Sigma F2016). The nuclei were identified by 4′,6′-diamidino-2-phenylindole staining. Nuclear import was analyzed with a Nikon Eclipse TE 300 microscope that has a fluorescence attachment and a Sony DKC-5000 charge-coupled device camera.

In-solution binding assays.

The interactions between GST-NLS_HPV16E6 and Kap α2, Kap β1, and Kap β2 were investigated using in-solution binding assays with the GST fusion proteins immobilized on glutathione-Sepharose (3 μg of protein/10 μl of beads) and the different recombinant Kaps in binding buffer (0.025% Tween and 0.5% glycerol in transport buffer). The final reaction volume was adjusted to 40 μl with binding buffer. In some experiments, the Kap βs were preincubated for 30 min with RanGTP (3 μg) before incubation with GST-NLS_HPV16E6. As a negative control we used GST, and as positive binding controls for Kap α2 we used GST-mpNLS_HPV16L1, and for Kap β2 we used GST-M9. After 60 min of incubation at room temperature, the beads were washed three times with binding buffer and the bound proteins were eluted with SDS sample buffer and analyzed by SDS-PAGE and Coomassie blue staining.

The HPV16 E6 oncoprotein can enter the nucleus in complex with either Kap β1 or Kap α2β1 heterodimers.

E6 oncoproteins interact with several nuclear transcription factors and coactivators in addition to cytoplasmic proteins and consequently have both nuclear and cytoplasmic activities. To identify and characterize the nuclear import pathway(s) exploited by the E6 oncoprotein of high-risk HPV16, we employed in vitro import assays in digitonin-permeabilized HeLa cells. Since recombinant E6 proteins are difficult to obtain in soluble form (43) and since GST fusion proteins have been extensively used in nuclear import studies, we used E6-GST fusion proteins that are readily obtained in soluble native form (26).

We have analyzed the nuclear import of a recombinant HPV16 E6-GST fusion protein and GST (as a control) in digitonin-permeabilized HeLa cells in the presence of only transport buffer or HeLa cytosol. We found that HPV16 E6-GST was transported into the nucleus in the presence of exogenous cytosol (Fig. 1B), but not in its absence (Fig. 1A). As expected, the GST reporter protein lacking an NLS was not imported into the nucleus in the presence of cytosol (data not shown). These data indicate that nuclear import of HPV16 E6 requires transport factors present in the cytosol.

It has been shown in transfection assays that deletion of the C-terminal region of HPV16 E6 (amino acids 121 to 151) abolishes its nuclear import (49), suggesting that this region functions as an NLS. Therefore, we investigated if the C-terminal region (amino acids 121 to 151) of HPV16 E6 is sufficient to target the GST reporter protein to the nucleus. The GST-E6_121-151 fusion protein was imported into the nuclei of digitonin-permeabilized cells in the presence of cytosol (Fig. 2B), whereas the GST control was not (Fig. 2D), confirming that, indeed, the C-terminal region of HPV16 E6 functions as an NLS.

Sequence analysis of HPV16 E6 NLS (₁₂₁KKQRFHNIRGRWTGRCMSCCRSSRTRRETQL₁₅₁) shows that it contains two blocks of basic amino acids separated by 19 amino acids. The first block of basic amino acids (₂₁KKQR) resembles a classic monopartite NLS and could interact with Kap α2 adapter, whereas the presence of many arginines throughout the NLS suggests the possibility of interaction directly with some members of the Kap β family (22).

To analyze the interactions between the NLS of HPV16 E6 with Kap α2 and Kap β1, we used in-solution binding assays. GST-NLS_HPV16E6, GST-NLS_HPV16L1 (positive control for Kap α2), and GST (negative control) were immobilized on glutathione-Sepharose beads and incubated with either Kap α2, Kap β1, or Kap α2 plus Kap β1. GST-NLS_HPV16E6 bound to both Kap α2 and Kap β1 and most likely formed a trimeric complex with Kap α2β1 heterodimers (Fig. 3A, lanes 1 to 3). As expected, GST-mpNLS_HPV16L1 bound to Kap α2 and formed a trimeric complex with Kap α2β1 heterodimers, but did not bind to Kap β1 (Fig. 3A, lanes 4 to 6). Mutation of R124 to G disrupts the first block of basic amino acids, resembling a monopartite NLS, and the corresponding NLS mutant would be expected to be unable to bind Kap α2. Indeed, the R124G NLS mutant did not bind Kap α2 (Fig. 3A, lane 7). Interestingly, the R124G NLS mutant still bound Kap β1, although the binding was reduced in comparison with the wild-type NLS (Fig. 3A, compare lanes 8 and 2). Neither Kap α2 nor Kap β1 bound to the immobilized GST (Fig. 3A, lanes 10 and 11). Significantly, binding of RanGTP to Kap β1 inhibited the interaction of GST-NLS_HPV16E6 with Kap β1 (Fig. 3B, compare lanes 1 and 2), suggesting that RanGTP can dissociate the HPV16 E6/Kap β1 complex.

Analysis of the nuclear import of HPV16 E6 in the presence of Kap β1 plus RanGDP or Kap α2 plus Kap β1 plus RanGDP revealed that, in agreement with the binding data, HPV16 E6 is imported into the nucleus in both cases (Fig. 4B and C). As expected, the control for the classical pathway, GST-mpNLS_HPV16L1, is imported only in the presence of Kap α2 plus Kap β1 plus RanGDP (Fig. 4I) and not in the absence of Kap α2 adapter (Fig. 4H). These data suggest that the HPV16 E6 oncoprotein can enter the nucleus via either the Kap β1-mediated pathway or the classical Kap α2β1-mediated pathway. Analysis of the nuclear import of the GST-E6_121-151 fusion protein in the presence of either Kap β1 or Kap α2 plus Kap β1 revealed that the NLS of HPV16 E6 is sufficient to mediate nuclear import via either the Kap β1- or the Kap α2β1-mediated pathway (Fig. 4E and F). Together the binding and nuclear import data suggest that the HPV16 E6 NLS can interact directly with either Kap α2 or Kap β1 and mediate entry into the nucleus via both Kap α2β1- and Kap β1-mediated pathways.

The HPV16 E6 oncoprotein interacts with Kap β2 and can enter the nucleus via a Kap β2-mediated pathway.

The presence of many arginines throughout the NLS of HPV16 E6 suggests the possibility of interaction directly with other members of the Kap β family (22). Therefore, we tested the ability of Kap β2 to interact with HPV16 E6 NLS using in-solution binding assays. GST-NLS_HPV16E6, GST-M9 (a positive control for Kap β2), and GST (negative control) immobilized on glutathione-Sepharose beads were incubated with Kap β2. We found that Kap β2 bound to GST-NLS_HPV16E6 and the GST-M9 positive control (Fig. 5A, lanes 1 and 2), but not to GST itself (lane 4). Interestingly, the R124G NLS mutant did not bind to Kap β2 (lane 3), suggesting that R₁₂₄ is essential for the interaction with Kap β2.

Moreover, binding of RanGTP to Kap β2 reduced the interaction of Kap β2 with either GST-NLS_HPV16E6 (Fig. 5B, compare lanes 1 and 2) or the GST-M9 control (data not shown). These binding data suggest that the HPV16 E6 oncoprotein could also enter into the nucleus via a Kap β2-mediated pathway. To test this, digitonin-permeabilized HeLa cells were incubated with either HPV16 E6-GST or M9-GST (as a positive control for Kap β2) in the presence of either transport buffer alone or RanGDP, or Kap β2 plus RanGDP. Both HPV16 E6 and M9-GST were imported into the nucleus in the presence of Kap β2 plus RanGDP (Fig. 6C and I). Moreover, GST-NLS_HPV16E6 was also imported into the nucleus in the presence of Kap β2 plus RanGDP (Fig. 6F). Together these data indicate that the HPV16 E6 oncoprotein interacts with Kap β2 via its C-terminal NLS and enters the nucleus via a Kap β2-mediated pathway.

An E6_R124G mutant has reduced nuclear import activity, and the E6_ΔKKQR deletion mutant is defective in nuclear import.

An HPV16 E6 mutant in which R₁₂₄ is changed to G has less ability to immortalize mammary epithelial cells than wild-type E6 (9). This mutation disrupted the ₁₂₁KKQR sequence, and the corresponding NLS mutant no longer bound Kap α2 and Kap β2 (Fig. 3A and 5A), suggesting that it would have reduced nuclear import activity. We analyzed the nuclear import of GST-NLS_R124G mutant in comparison with the GST-NLS_E6 wild type in identical conditions in digitonin-permeabilized cells in the presence of cytosol. Quantitation of nuclear import of the GST-NLS_R124G mutant and the GST-NLS_E6 wild type was carried out by measuring the mean nuclear fluorescence using IPLAB software. We found that nuclear import of the GST-NLS_R124G mutant was reduced in comparison with that of the GST-NLS_E6 wild type (Fig. 7, compare panels A and B), with a mean ratio of mutant nuclear import/wild-type nuclear import of 0.58 ± 0.05 (in four experiments).

We also compared the nuclear import of wild-type HPV16 E6 with that of two full-length E6 mutant proteins, one containing the R124G mutation and the other containing a deletion of the first block of basic amino acids (KKQR) of the E6 NLS. Nuclear import of the E6_R124G mutant was reduced in comparison with that of wild-type E6 (Fig. 8, compare panels A and B), with a mean ratio of mutant nuclear import/wild-type nuclear import of 0.75 ± 0.02 (in 4 experiments). Significantly, the E6_ΔKKQR deletion mutant was not imported into the nucleus (Fig. 8C). These data suggest that the ₁₂₁KKQR₁₂₄ sequence within the NLS of the HPV16 E6 oncoprotein is essential for E6 nuclear import.