Nuclear Export of Human Papillomavirus Type 31 E1 Is Regulated by Cdk2 Phosphorylation and Required for Viral Genome Maintenance

Plasmid constructions and mutagenesis.Plasmids expressing dominant-negative (dn) HA-Cdk2, 6xmyc-cyclin E, 6xmyc-cyclin A, and p21-6xhis proteins were kindly provided by Sylvain Meloche (University of Montreal) and were described previously (8, 40). The plasmid used to express EYFP-31E1 was previously described (7). In this plasmid, the splicing site of HPV31 E1 (GCAGGT) was changed to GCTGGC by mutagenesis using the following primer: 5′-GCTGATCCAGCTGGCACAGATGGGGAGGGGACGG-3′. EYFP-31E1 expression plasmids, as well as HPV31 genomes harboring mutation of either the cyclin-binding motif (CBM; R123A, R124A, and L125A), the NLS (K86G and R87G), the nuclear export signal (NES; L109A and I112A), or substitution of the five putative Cdk2 phosphorylation sites to alanine (S92, S106, T169, T316, and T506) or aspartate (S92 and S106) were constructed by site-directed mutagenesis using the QuikChange mutagenesis kit (Stratagene). All DNA constructs were verified by sequencing. Further details on their construction will be made available upon request.

Cell culture and transfection.The human cervical carcinoma cell line C33A was grown in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 0.5 IU of penicillin/ml, 50 μg of streptomycin/ml, and 2 mM l-glutamine. Human foreskin keratinocytes (HFKs) were maintained in KGM (Clonetics) or in E medium in the presence of mitomycin C (Boehringer Mannheim)-treated fibroblast feeders. Transfections of C33A cells were performed using the Lipofectamine 2000 reagent (Invitrogen), and HFKs were transfected using the Fugene 6 reagent (Roche) according to the manufacturer's protocol.

Confocal fluorescence microscopy.Approximately 8 × 10⁵ C33A cells were transfected with 0.5 μg of EYFP-31E1 expression plasmid and grown on coverslips. When indicated, 0.5 μg of 3F-31E2 (WT or E39Q) or 1.0 to 2.0 μg of an expression plasmid for the dominant-negative form of Cdk2 (Cdk2dn) or p21 were cotransfected with EYFP-31E1. At 24 h posttransfection, cells were fixed with 4% formaldehyde and permeabilized with 0.2% Triton X-100. DNA was stained with TO-PRO-3 (Molecular Probes). Cells were mounted using Vectashield mounting medium (Vector Laboratories). When indicated, leptomycin B (LMB [Sigma-Aldrich]; catalog no. L2913) (15 μg/ml) was added to the cells 6 h before fixation. Images were acquired by using a LSM510 confocal laser coupled to an Axiovert 100M inverted scanning microscope (Zeiss, Toronto, Canada) and analyzed using LSM Image Browser version 3.2.0.70 (Zeiss).

GST fusion proteins and pulldown assays.PCR-amplified E1 fragments encompassing either the OBD alone (amino acids [aa] 191 to 353 for HPV11 and aa 170 to 332 for HPV31) or the OBD together with the N-terminal region of E1 (aa 1 to 353 for HPV11 and aa 1 to 332 for HPV31) were inserted between the BamHI and EcoRI sites of the plasmid AB-401, a modified version of pGEX-4T-1 (GE Healthcare) in which a sequence encoding a hexahistidine tag was inserted downstream of the glutathione S-transferase (GST) coding region. GST fusion proteins were expressed in Escherichia coli BL21(DE3) (Novagen) and purified as previously described (46). Protein concentrations were determined using the Bio-Rad Bradford analysis. GST pulldown assays were performed as described previously (7, 46).

Immunoprecipitations.For coimmunoprecipitation studies, approximately 4 × 10⁶ C33A cells were transfected with 3 μg of EYFP-31E1 expression plasmid along with 3 μg of 6xmyc-tagged cyclin E or A expression plasmid in a 100-mm plate. Transfected cells were lysed 48 h posttransfection in coimmunoprecipitation buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, protease and phosphatase inhibitors [10 μg of antipain, 2 μg of leupeptin, 1 μg of pepstatin, and 2 μg of aprotinin/ml; 1 mM phenylmethylsulfonyl fluoride; 50 mM NaF; and 1 mM orthovanadate {Na₃VO₄}]) for 30 min at 4°C. Cleared whole-cell extracts were incubated with 1 μg of anti-GFP antibodies coupled with protein G-Sepharose (GE Healthcare) for 180 min. The beads were washed three times with TBS buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl), and the bound proteins were analyzed by SDS-PAGE and Western blotting. For the detection of phosphorylated E1, EYFP-31E1 proteins were immunoprecipitated essentially as described above with the exception that the transfected cells were lysed 24 h posttransfection in immunoprecipitation buffer (i.e., coimmunoprecipitation buffer supplemented with 10 mM iodoacetamide, 20 mM N-ethylmaleimide [NEM], and 20 μM MG132) and that the beads were washed with TBS buffer supplemented with protease and phosphatase inhibitors (10 μg of antipain, 2 μg of leupeptin, 1 μg of pepstatin, and 2 μg of aprotinin/ml; 1 mM phenylmethylsulfonyl fluoride; 50 mM NaF; 1 mM orthovanadate [Na₃VO₄], 10 mM iodoacetamide; 20 mM NEM; 20 μM MG132).

Antibodies and Western blotting.Endogenous Cdk2 and cyclin A were detected using mouse monoclonal antibody from BD Transduction Laboratory (catalog no. 610145) and Lab Vision Corp. (catalog no. Ab-6), respectively. The mouse monoclonal antibody against the myc epitope that was used to detect the 6xmyc-tagged cyclin E and cyclin A proteins was purchased from Santa Cruz (catalog no. sc-40). Green fluorescent protein (GFP) fusion proteins were detected using a mixture of two mouse monoclonal antibodies purchased from Roche (catalog no. 11814460001), while β-tubulin was detected using a mouse monoclonal antibody from Sigma-Aldrich (catalog no. T0426). Phosphorylation of EYFP-31E1 was detected using a mouse monoclonal anti-phospho-serine/threonine-proline (P-Ser/Thr-Pro, MPM-2) from Millipore (catalog no. 05-368). For Western blot analysis, proteins were transferred onto polyvinylidene difluoride membranes and detected using horseradish peroxidase-conjugated sheep anti-mouse secondary antibody from GE Healthcare (catalog no. NA931) and an enhanced chemiluminescence detection kit (GE Healthcare).

Transient HPV DNA replication assay.Transient HPV31 DNA replication was performed as described previously (13) but using EYFP-31E1 instead of Flag-tagged E1 (3F-31E1). Briefly, ∼5 × 10⁴ C33A cells were transfected with four plasmids encoding, respectively, EYFP-31E1, 3F-31E2, the minimal origin of DNA replication together with a firefly luciferase reporter gene, and Renilla luciferase. Replication of the origin-containing plasmid was quantified by measuring the levels of firefly and Renilla luciferase activities using the Dual-Glo luciferase assay system (Promega) at 72 h posttransfection. Each E1 mutant was analyzed in duplicates, in two separate experiments. The results are reported as the mean of the four independent values, and error bars represent the standard deviation.

HPV31 genome maintenance and amplification assays.WT and mutant HPV31 genomes (accession no. J04353 [18]) were released from the pBR322 min plasmid by restriction enzyme digestion, followed by unimolecular ligation with T4 DNA ligase (New England Biolabs), as described previously (25). The genomes were then precipitated in 35% isopropyl alcohol and 10% NaCl and resuspended in TE (10 mM Tris-HCl, 1 mM EDTA [pH 7.5]). Genomes were cotransfected with pSV₂neo vector in HFKs grown to 30% confluence in 10-cm dishes. After 24 h, the transfected cells were treated with trypsin and replated onto tissue culture dishes containing E medium with epidermal growth factor and J2 fibroblast feeder cells. G418 selection was done as previously described (11). Selected HFKs were resuspended in 1.5% methylcellulose-containing E medium (methylcellulose solution was prepared as previously described [11]) to induce differentiation. Fibroblast feeders were removed prior to cell harvesting at 0, 24, and 48 h by a 2-min treatment with phosphate-buffered saline containing 0.5 mM EDTA. After centrifugation and washing, the cells were resuspended in DNA lysis buffer (400 mM NaCl, 10 mM Tris-HCl [pH 7.4], 10 mM EDTA, 50 μg of RNase A/ml, and 0.2% SDS), followed by incubation overnight at 37°C with proteinase K (50 μg/ml). Samples were then passed through an 18-gauge needle 10 times to shear the DNA, which was next extracted by phenol-chloroform, followed by ethanol precipitation. TE-resuspended DNA was finally analyzed by Southern blotting as previously described (54) to detect HPV31 genomic DNA.

Flow cytometry analysis.Approximately 2 × 10⁶ C33A cells were transfected with 3 μg of the indicated plasmid(s) in 60-mm plates. At 4 h posttransfection, the cells were treated with trypsin and seeded in 60-mm plates at a low density (5 × 10⁵ cells/plate) to promote cell division. After 24 h, the culture medium was changed and replaced with either fresh medium only (asynchronous population) or with fresh medium containing 500 μM mimosine to synchronize cells in G₁/S. When indicated, cells were released from the mimosine block into fresh medium containing 0.5 μg of nocodazole/ml to block cell cycle progression in G₂/M for a period of 24 h. For cell cycle analysis, cells were treated with trypsin, resuspended at a concentration of 10⁶ cells/ml, and stained using fresh medium supplemented with 1 μg of Hoechst (Sigma catalog no. B2261)/ml and 1 μg of Verapamil (Sigma catalog no. V4629)/ml for 30 min at 37°C. The DNA content of enhanced yellow fluorescent protein (EYFP)-expressing cells was determined by flow cytometry on a FACS BD LSR flow cytometer using the CellQuest Pro Software (BD). The cell cycle distribution was further analyzed and quantified using the FlowJo (v8.1) and ModFit LT softwares, respectively.

Colony formation assay.Approximately 1.2 × 10⁶ C33A cells were transfected with 1.5 μg of the indicated plasmid in a six-well plate. At 24 h posttransfection, cells were treated with trypsin and seeded onto 100-mm plates in fresh medium supplemented with 500 μg of G418/ml. Antibiotic selection was maintained for a period of approximately 3 weeks. Cells were then fixed for 10 min in cold methanol and stained for 2 min with methylene blue (1% [wt/vol] in 60% MeOH-H₂O).

A conserved CBM in HPV31 E1 inhibits its CRM1-dependent nuclear export.HPV11 E1 has been shown to shuttle between the nucleus and the cytoplasm through a bi-partite NLS and a leucine-rich NES, both of which are highly conserved (Fig. 2A). To verify whether these sequences are also mediating the shuttling of HPV31 E1, we mutated them in the context of EYFP-31E1 and examined the localization of the resulting mutant proteins in C33A cells by confocal fluorescence microscopy. As shown in Fig. 2B, WT E1 was found almost exclusively in the nucleus of transfected cells. Examination of 300 EYFP-positive cells allowed us to quantify this result (Table 1) and determine that E1 was exclusively nuclear (N) in 95% of cells while being located in both the nucleus and the cytoplasm (NC) in the remaining 5%. As expected, mutation of the NLS completely abolished the exclusively nuclear accumulation of E1 (Fig. 2B, N = 0%). The E1 NLS mutant was found either all in the cytoplasm (C = 42%) or in both the nucleus and the cytoplasm (NC = 58%). In contrast and as expected, the E1 NES mutant protein was as nuclear as WT E1 (Fig. 2B, N = 97%). Because the nuclear export of HPV11 E1 was shown to be inhibited by cyclin E/A-Cdk2 (10), which binds directly to E1 through a conserved RxL CBM (34), we investigated the effect of a CBM mutation on the localization of EYFP-31E1. First, we determined by GST pulldown assay that the N-terminal region of HPV31 E1 is able to interact with Cdk2 as efficiently as that of HPV11 E1 (Fig. 2C). Next, we confirmed by coimmunoprecipitation that HPV31 E1 can interact with both cyclin E and cyclin A through its CBM (Fig. 2D). We then examined the intracellular localization of the EYFP-31E1 CBM mutant protein and found that a substantial portion of the protein was localized to the cytoplasm (Fig. 2B, NC = 89%, C = 7%). Furthermore, we found that this cytoplasmic accumulation could be abrogated by mutation of the NES (CBM/NES double mutant; Fig. 2B, N = 95%), suggesting that the cytoplasmic accumulation of the E1 CBM mutant results from its nuclear export. In further support of this suggestion, we found that the cytoplasmic accumulation of the E1 CBM mutant could be prevented by LMB, an inhibitor of the CRM1-dependent nuclear export pathway (12, 16, 29). As a control, we verified by Western blotting that all of the mutant EYFP-31E1 proteins described above were expressed at comparable levels as WT E1 (Fig. 2E). The results presented thus far suggested that the interaction of HPV31 E1 with cyclin E/A-Cdk2 promotes its nuclear accumulation by inhibiting its nuclear export. To determine whether the kinase activity of Cdk2 is required to prevent E1 nuclear export, we tested the effect of a dominant-negative Cdk2 (Cdk2dn) and of the endogenous Cdk2 inhibitor p21 (20). As anticipated, inhibition of Cdk2 activity resulted in the relocalization of EYFP-31E1 from the nucleus to the cytoplasm (Fig. 2B and Table 1). Collectively, these results demonstrate that HPV31 E1 shuttles between the nucleus and the cytoplasm using the same regulatory elements (NLS, NES, and CBM) and cellular factors (cyclin E/A-Cdk2 and CRM1) as HPV11 E1.

• Open in new tab
• Download powerpoint

FIG. 2.

Nuclear export of HPV31 E1 is regulated by Cdk2 activity. (A) Alignment of the N-terminal regions from different papillomavirus E1 encompassing the bipartite nuclear NLS, the NES, and the CBM. (B) Intracellular localization of WT and mutant HPV31 E1. C33A cells transiently expressing the WT or indicated mutant EYFP-31E1 proteins were visualized by fluorescence confocal microscopy. Nuclei (DNA) were stained with TO-PRO-3. When indicated, cells were also transfected with Cdk2dn or p21 to inhibit Cdk2 activity or were treated with 15 μg of LMB/ml to inhibit CRM1-dependent nuclear export. Quantification of these results is presented in Table 1. (C) Interaction of HPV31 E1 with cyclin A and Cdk2 in vitro. The indicated HPV11 and HPV31 E1 fragments were purified as GST fusion proteins and used in pulldown assays with EcR-293 whole-cell extracts. Bound proteins were separated by SDS-PAGE and analyzed by Western blotting with anti-cyclin A (CycA) and anti-Cdk2 antibodies, as indicated. The bottom panel shows a Coomassie blue-stained SDS-PAGE gel of the purified GST-E1 proteins used in these pulldown assays. The sizes of molecular weight markers are indicated on the left of the gel. (D) Interaction of HPV31 E1 with cyclin E and A in vivo is mediated by the CBM motif. EYFP-31E1 WT (W) or CBM (C), NES (N), or NLS (n) E1 mutants were cotransfected with cyclin E (Cyc E) or cyclin A (Cyc A) fused to a 6xmyc epitope (myc). Coimmunoprecipitated proteins were separated by SDS-PAGE and analyzed by Western blotting with anti-GFP and anti-myc antibodies, as indicated. The lower panels show the amount of EYFP proteins expressed in the different whole-cell extracts prior to their immunoprecipitation (input). Tubulin was used as a loading control (α-Tub.). (E) Western blot analysis of total protein extracts prepared from transfected C33A cells expressing either WT or the indicated EYFP-31E1 mutant proteins. Extracts were prepared 24 h posttransfection. E1 proteins were detected by using an anti-GFP antibody (α-GFP), and tubulin was used as a loading control (α-Tub.).

Serines 92 and 106 of HPV31 E1 are essential for its nuclear accumulation.The results presented above suggested that E1 needs to be phosphorylated by Cdk2 to prevent its nuclear export (i.e., to accumulate in the nucleus). HPV31 contains five putative Cdk2 phosphorylation sites (S/T-P): S92, S106, T169, T316, and T506. To test whether any one of these residues is essential for nuclear accumulation of EYFP-31E1, we mutated them to alanine, either individually or in combination, and determined the intracellular localization of the resulting mutant proteins in C33A cells. We found that two mutants, S92A and S106A, were partially relocalized to the cytoplasm (Fig. 3A and Table 1). Given that the relocalization of these two single mutants was not as robust as that observed with the CBM mutant (Fig. 2B), we constructed a double mutant, S92A/S106A (SSAA), and found that it recapitulated completely the cytoplasmic localization of the CBM mutant (Fig. 3A and Table 1). Furthermore, cytoplasmic accumulation of the double mutant was found to be the result of its nuclear export as it was inhibited by LMB. None of the three other putative Cdk2 sites, either alone or in pairs, affected the nuclear accumulation of the protein (Fig. 3A and data not shown). To definitively rule out the involvement of these other sites (T169, T316, and T506), we created a mutant protein in which all five Cdk2 sites were mutated simultaneously to alanine (AAAAA). As expected, this quintuple mutant was as cytoplasmic as the CBM or S92A/S106A (SSAA) mutant proteins (Fig. 3B and Table 1). To determine whether S92 and S106 are sufficient to prevent nuclear export of E1, we reintroduced them in the AAAAA mutant protein, either alone or in combination. Satisfyingly, both single serine mutants (SAAAA and ASAAA) accumulated mainly in the nucleus, and the double serine mutant (SSAAA) fully recapitulated the nuclear accumulation observed with WT E1 (Fig. 3B, Table 1). We did not observe any significant effect of the aforementioned mutations on the expression of EYFP-31E1 in repeated Western blotting experiments (Fig. 3C and data not shown). Collectively, the results presented above indicated that S92 and S106 are the only two putative Cdk2 sites that are necessary and sufficient to prevent nuclear export of HPV31 E1.

• Open in new tab
• Download powerpoint

FIG. 3.

Serines 92 and 106 of HPV31 E1 are essential for its nuclear accumulation. Intracellular localization of EYFP-31E1 proteins in which one, two (A) or several (B) putative Cdk2 phosphorylation sites have been substituted for alanine. AAAAA refers to a HPV31 E1 mutant in which all five putative Cdk2 phosphorylation sites have been substituted for alanine, as described in the text. Where indicated, cells were treated with 15 μg of LMB/ml to inhibit nuclear export. Quantification of these results is presented in Table 1. (C) Western blot analysis of C33A cells expressing WT or the indicated EYFP-31E1 mutant proteins as described in Fig. 2E. E1 proteins were detected by using an anti-GFP antibody (α-GFP), and tubulin was used as a loading control (α-Tub.).

Evidence that HPV31 E1 is phosphorylated on serines 92 and 106 by Cdk2.The results presented above suggest that Cdk2 phosphorylation of E1 on S92 and S106 is required to prevent its nuclear export, although they did not show directly that these two serines are phosphorylated. To determine whether S92 and S96 are indeed phosphorylated by Cdk2 in vivo, we immunoprecipitated EYFP-E1 from transfected C33A cells with an anti-GFP antibody and analyzed the precipitate by Western blotting with an anti-phospho-Ser/Thr-Pro antibody. As specificity controls, we used the aforementioned E1 protein in which all five putative Cdk2 sites were mutated to alanine (AAAAA mutant), as well as derivatives in which only S92 (SAAAA), S106 (ASAAA), or both serines (SSAAA) were left intact. Figure 4A shows that the immunoprecipitated E1 reacted with the anti-phospho-Ser/Thr-Pro antibody, whereas the quintuple alanine mutant did not, thus indicating that E1 is phosphorylated in vivo on one or more of these five sites. Interestingly, we found that E1 was only poorly phosphorylated when only S92 or S106 was present (SAAAA or ASAAA) but phosphorylated near WT levels when both of these serines were present together (SSAAA). These results are in excellent agreement with our finding presented above that both serines are required to efficiently prevent the nuclear export of E1 (Fig. 3). To ascertain that the phosphorylation of E1 was due to Cdk2, we performed similar experiments from cells cotransfected with expression vectors for cyclin E and Cdk2 to boost the intracellular levels of this kinase or, conversely, from cells cotransfected with p21 to inhibit Cdk2. As anticipated, we found that the levels of E1 phosphorylation were increased in cells overproducing cyclin E-Cdk2 (Fig. 4B) and greatly reduced in cells overproducing p21 (Fig. 4C). Overproduction of cyclin E-Cdk2 led to an increase in the phosphorylation of both S92 and S106 individually (SAAAA and ASAAA) (Fig. 4B). Significantly, maximal phosphorylation of E1 was achieved when both serines were present together (SSAAA; Fig. 4B), similarly to what was observed in the absence of cyclin E-Cdk2 overproduction. Collectively, these results provide strong evidence that E1 is phosphorylated in vivo by Cdk2 on S92 and S106 and that both residues are needed for efficient phosphorylation of E1.

• Open in new tab
• Download powerpoint

FIG. 4.

Evidence that HPV31 E1 is phosphorylated on serines 92 and 106 by Cdk2. (A) Western blots of immunoprecipitated EYFP-31E1 fusion proteins, or of EYFP alone as a control, using an anti-phospho-Ser/Thr-Pro antibody (upper panel). Because these proteins were immunoprecipitated with an anti-GFP antibody, an anti-GFP Western blot was also performed to control for the efficiency of immunoprecipitation (lower panel). WT E1 (WT) or mutant derivatives in which all five putative Cdk2 sites were mutated to alanine (AAAAA) or in which only S92 (SAAAA), S106 (ASAAA), or both serines (SSAAA) were left intact were used. (B and C) Same as in panel A, but using EYFP-31E1 fusion protein immunoprecipitated from C33A cells cotransfected with expression vectors for cyclin E and Cdk2 (B, +Cyc E-Cdk2) or for p21 (C, +p21). The positions of EYFP, EYFP-31E1, and of the anti-GFP antibody heavy (H) and light (L) chains are indicated.

Cytoplasmic E1 mutants can be relocalized to the nucleus by E2 and are able to support transient viral DNA replication.The intracellular localization studies presented above were all performed with E1 alone, in the absence of E2. Because E2 interacts with E1 and is itself a nuclear protein, we investigated whether it could relocalize cytoplasmic E1 mutants to the nucleus. First, we tested the effect of E2 on the localization of the EYFP-31E1 NLS mutant. As shown in Fig. 5A, E2 was not only able to relocalize the E1 NLS mutant from the cytoplasm to the nucleus but also promoted its accumulation into discrete nuclear foci similarly to what is observed with WT E1 and E2 (Fig. 5A and Table 2). E2 has been shown to accumulate in nuclear foci whose function remains unclear, although it has been suggested that they may be precursors to DNA replication foci (53). Regardless of their exact function, E2 was able to recruit E1 to these foci in a manner that was dependent on the interaction between both proteins, as an E2 mutant with a reduced affinity for E1 (E2 E39Q) (1, 13, 42, 47) was unable to relocalize E1 efficiently (Fig. 5A). Next, we tested whether E2 was also able to promote the accumulation of different E1 phosphorylation site mutants into nuclear foci. This was indeed the case for all mutants tested, including the one lacking the CBM and those lacking S92 and/or S106 (Fig. 5A and Table 2).

• Open in new tab
• Download powerpoint

FIG. 5.

Cytoplasmic E1 proteins are relocalized to the nucleus by E2 and are able to support transient viral DNA replication. (A) Intracellular localization of WT or the indicated mutant EYFP-31E1 proteins coexpressed with either WT (left panel) or E39Q (right panel) HPV31 E2. Quantification of these results is presented in Table 2. (B) Transient DNA replication activities of the indicated EYFP-31E1 proteins in C33A cells. DNA replication activities were measured by determining the ratio of firefly (Fluc) to Renilla (Rluc) luciferase activity as described in Materials and Methods. Cells transfected with EYFP only (No E1) were used as a negative control. Replication activities are reported as a percentage of the Fluc/Rluc ratio obtained with the highest amount of WT E1 expression vector.

These results prompted us to test whether these E1 mutants could support transient viral DNA replications, using our recently developed HPV31 assay in which replication of an ori-containing plasmid by E1 and E2 results in increased expression of a firefly luciferase gene located in cis of the ori (13). For each E1 mutant, three different amounts of expression plasmid were used, along with a constant amount of E2 and of the origin. A Renilla luciferase plasmid was also cotransfected for normalization, as previously described (13). Under these conditions, the E1 NLS mutant was able to support ca. 70% of the WT level of DNA replication. A similar finding was obtained with the CBM mutant. Interestingly, we found that the addition of the NES mutation in the CBM mutant (to create a CBM/NES double mutant) was sufficient to restore almost WT levels of DNA replication (Fig. 5B). This result demonstrated that the CBM, and we surmise the interaction of E1 with cyclin E/A-Cdk2, is not essential for viral DNA replication per se, as long as the E1 CBM mutant is maintained in the nucleus in this case by mutation of its NES. Consistent with the interaction of E1 with cyclin E/A-Cdk2 being dispensable for viral DNA replication, we found that none of the putative Cdk2 site mutants were replication defective, including the S92A/S106A double mutant (SSAA, Fig. 5B). Together, these results and those presented in the previous sections indicate that the CBM, S92, and S106 are not essential for viral DNA replication per se but rather are needed to facilitate nuclear accumulation of E1 in the absence of E2. Although the inhibition of E1 nuclear export by cyclin E/A-Cdk2 is not essential for viral DNA replication in transfected cells because E2 can relocalize cytoplasmic E1 to the nucleus, it is likely an important process during a normal infection where the levels of E2 are lower and likely insufficient to counteract the nuclear export of E1.

Nuclear export of HPV31 E1 is inhibited by phosphomimetic substitution of S92 and S106 and is dispensable for transient viral DNA replication.The results presented above suggested that phosphorylation of S92 and S106 is necessary and sufficient to prevent E1 nuclear export. To test this possibility more directly, we investigated whether the substitution of S92 and/or S106 by aspartate, as a phosphomimetic, in EYFP-31E1 was sufficient to inhibit its nuclear export. We found that all three E1 proteins accumulated in the nucleus to a similar extent as the WT protein (S92D, N = 96%; S106D, N = 94%; SSDD, N = 97%) (Fig. 6A and Table 1). Presumably, nuclear accumulation of these phosphomimetic E1 mutants is now independent of Cdk2 activity. To verify this hypothesis, we inhibited Cdk2 by transfection of its inhibitor p21, which we showed in Fig. 2B promotes the nuclear export of WT E1. As expected, whereas a high proportion of WT E1 was relocalized to the cytoplasm by the action of p21, all three phosphomimetic mutants remained nuclear, similarly to the E1 NES mutant, which was used as a control (Fig. 6A and Table 1). We also verified by Western blotting that all three phosphomimetic E1 mutants were expressed at similar levels as WT E1 (Fig. 6B).

• Open in new tab
• Download powerpoint

FIG. 6.

Nuclear export of HPV31 E1 is inhibited by phosphomimetic substitution of S92 and S106 and is dispensable for transient viral DNA replication. (A) Intracellular localization of EYFP-31E1 and of mutant derivatives carrying a mutation in the NES or in which S92 and/or S106 have been changed to an aspartic acid as a phosphomimetic. In the right panel, the same proteins were cotransfected with the Cdk2 inhibitor p21. Quantification of these results is presented in Table 1. (B) Western blot analysis of C33A cells expressing WT or the indicated EYFP-31E1 mutant proteins as described in Fig. 2E. E1 proteins were detected by using an anti-GFP antibody (α-GFP), and tubulin was used as a loading control (α-Tub.). (C) Transient DNA replication activities of the indicated EYFP-31E1 proteins in C33A cells. DNA replication activities were measured by determining the ratio of firefly (Fluc) to Renilla (Rluc) luciferase activity as described in Materials and Methods. Cells transfected with EYFP only (No E1) were used as a negative control. Replication activities are reported as a percentage of the Fluc/Rluc ratio obtained with the highest amount of WT E1 expression vector.

Next, we used these E1 mutants that are confined to the nucleus to test whether nuclear export of E1 is required for transient viral DNA replication. Specifically, we tested the activities of the NES, S92D, S106D, and S92D/S106D E1 mutants in our luciferase DNA replication assay, using three different concentrations of expression vector for each E1 protein. These experiments revealed that all four E1 mutants were as active as the WT protein (Fig. 6C). In fact, we consistently observed slightly higher levels of replication with these E1 mutants that are confined to the nucleus. Collectively, the results presented above indicated that phosphomimetic mutations at S92 and S106 are sufficient to prevent export of E1 to the cytoplasm, even in the absence of Cdk2 activity. Furthermore, they showed that transient DNA replication is independent of E1 nuclear export, at least under our assay conditions.

Nuclear export of HPV31 E1 is required for viral genome maintenance in HFKs.The aforementioned results showed that nuclear export of HPV31 E1 is highly regulated but not required for its DNA replication function per se. This prompted us to investigate whether nuclear export of E1 is required for maintenance and amplification of the viral genome, two important stages of the HPV life cycle that rely on DNA replication. To do this, we used a HPV31 genome carrying the E1 NES mutation, as well as the WT genome as a control, to immortalize primary HFKs, as described previously (59). Southern blotting analysis was then performed to assess whether the NES genome could be maintained in episomal form and, if so, whether it could be amplified upon cellular differentiation (27, 41, 59). We found that in contrast to the WT genome, only low levels of episomes were detected for the NES mutant in HFKs from two independent donors (Fig. 7A, compare 0 h in methylcellulose for WT and NES). Interestingly, the low level of mutant episome detected in one of the donors was increased upon cellular differentiation induced by growth of the HFKs in methylcellulose (Fig. 7A), indicating that the NES mutant genome could be amplified and thus that the E1 NES mutant helicase is functional. Collectively, these results indicated that the primary defect of the NES genome is a failure to be properly maintained in episomal form in undifferentiated cells rather than a defect in late functions. We further characterized this phenotype by monitoring the amount of mutant episome over several cell passages. These studies indicated that the NES episome was progressively lost upon cell division, in contrast to the WT, which remained at a constant level throughout the experiment (Fig. 7B). To rule out that this maintenance phenotype was due to a nonspecific effect of the NES mutation, we also characterized a genome carrying the E1 SSDD mutation (in which serines 92 and 106 were both substituted by aspartate as phosphomimetics). Similar to what was observed with the NES genome, the E1 SSDD mutant was not maintained efficiently as an episome (Fig. 7C). Altogether, the results presented above indicated that nuclear export of E1 is essential for proper maintenance of the episome in undifferentiated cells but not for its amplification upon differentiation.

• Open in new tab
• Download powerpoint

FIG. 7.

Nuclear export of HPV31 E1 is required for genome maintenance in HFKs. (A) Maintenance and amplification of HPV31 genomes expressing either the WT or NES mutant E1 protein. HFKs immortalized with the WT or the NES genome were resuspended in methylcellulose (MC) for 24 and 48 h to induce their differentiation. DNA was harvested from undifferentiated (0 h) and from differentiated cells (24 and 48 h). The figure shows the Southern blots used to detect the presence of episomal genomes. Two independent experiments are presented that were performed using HFKs from two different donors (d1 and d2). (B) Southern blots showing the maintenance of the HPV31 WT or NES genome in immortalized HFKs at the indicated number of cell passages. (C) Maintenance and amplification of an HPV31 genome expressing the S92D/S106D (SSDD) E1 mutant in HFKs from donor d2. Experiments were performed as described in panel A.

E1 interferes with cell cycle progression.One possibility to explain the inability of the E1 NES genome to be maintained in episomal form would be that it is poorly replicated and, as a result, lost upon cell division. However, we did not favor this possibility given that we had observed that the E1 NES mutant protein is competent for transient DNA replication (Fig. 6C) and for genome amplification (Fig. 7A). Instead, we considered the possibility that the NES genome has a deleterious effect on cell proliferation, perhaps caused by sustained or unregulated viral DNA replication, and thus selected against over multiple cell passages. One reason for favoring this possibility was our fortuitous observation that cells undergoing transient HPV DNA replication are altered in their cell cycle progression. Indeed, cell cycle analysis of C33A cells expressing EYFP-31E1 and 3F-31E2 proteins and containing the Ori plasmid (DNA replication conditions) revealed that the proportion of EYFP-expressing cells that are in S phase is increased by ca. 15% compared to mock-transfected EYFP-expressing cells (left panels in Fig. 8A, Table 3). As would be expected from a defect in cell cycle progression, this increase in S-phase cells was accompanied by a reduction of cells in G₁ and G₂/M. Similar findings were obtained when these experiments were repeated with the E1 NES mutant protein (Fig. 8A).

• Open in new tab
• Download powerpoint

FIG. 8.

HPV31 E1 impairs cell cycle progression in S-phase. (A and B) C33A cells were transfected with the indicated EYFP-31E1 or EYFP expression vector either alone or together with an expression vector for 3F-31E2 and the ori-containing plasmid. After 48 h, cells were treated with trypsin, and their DNA was stained with Hoechst. The DNA content of EYFP-expressing cells was then analyzed by flow cytometry. For each transfection condition described on the left side of the figure, the cells were either grown asynchronously, synchronized in G₁/S with mimosine (T = 0) or synchronized with mimosine for 24 h and then released in nocodazole for an additional 24 h (T = 24 h), as indicated. The cell cycle distribution of each sample was derived from the analysis of 5,000 EYFP-expressing cells and is represented by a histogram. For the analysis of untransfected cells (Mock), 5,000 Hoechst-positive cells were collected. Quantification of these results is reported in Table 3.

To distinguish whether these cells were blocked in S phase or only progressing very slowly through S phase, we synchronized them in G₁/S with mimosine and determined their capacity to resume proliferation after removal of the inhibitor. Specifically, we treated cells for 24 h with mimosine, verified that they were efficiently blocked in G₁/S (center panels in Fig. 8A), and then released them in medium containing nocodazole for an additional 24 h to allow their progression to G₂/M (right panels in Fig. 8A). This analysis showed that while 80% of the control cells expressing EYFP alone were able to progress beyond the early S phase, only 36% of cells expressing either WT or the NES mutant EYFP-31E1 (together with E2 and the ori) were able to do so (Fig. 8A). These results suggested that cells undergoing transient DNA replication are not completely blocked but rather delayed in their progression through early S phase.

To probe the mechanism underlying this cell cycle delay, we first tested whether DNA replication itself was responsible for this effect by repeating the experiment with an E2 mutant (E39Q) that does not support DNA replication (13). Cells expressing E2 E39Q (together with EYFP-31E1 and ori) showed an altered cell cycle profile similar to that obtained with WT E2, EYFP-31E1, and ori (Fig. 8B), indicating that DNA replication per se is not required for this effect. In contrast, in cells in which EYFP-31E1 was omitted and replaced by EYFP alone, no cell cycle delay was observed (Fig. 8B). This pointed to E1 as being responsible for slowing down S-phase progression. Therefore, we tested directly whether E1 alone could cause a cell cycle delay. We observed that EYFP-31E1, either WT or containing the NES mutation, was sufficient to impair cell cycle progression (Fig. 8B). This effect was specific to E1 since EYFP-31E2 and EYFP alone had no effect under the same conditions (Fig. 8B). Collectively, these results showed that HPV31 E1 impairs cell cycle progression independently of its capacity to interact with E2 or the origin.

Nuclear export of E1 alleviates its detrimental effect on S-phase progression.The finding that expression of E1 impairs cell cycle progression (Fig. 8B) prompted us to investigate whether nuclear export of E1 could attenuate this phenotype. Specifically, we investigated the effect of different E1 mutants on the cell cycle distribution of asynchronous cells (left panels in Fig. 9A) and, more importantly, on the ability of these E1-expressing cells to progress beyond early S phase following synchronization with mimosine and release in nocodazole-containing medium, as described above (right panels in Fig. 9A). Under these conditions, 45% of the cells expressing WT EYFP-31E1 were able to progress to G₂/M, whereas 80% of cells expressing EYFP alone were able to do so (Table 3). As expected, similar results were obtained with the E1 NES mutant protein that is confined to the nucleus (Fig. 9A). Next, we tested the effect of the E1 CBM mutant protein that is constitutively exported to the cytoplasm and found that it had little to no effect on S-phase progression (Fig. 9A). Thus, nuclear export of E1 prevents its deleterious on cell cycle progression. As expected, mutation of the NES in the context of the E1 CBM mutant to prevent its nuclear export (CBM/NES double mutant) recapitulated the S-phase delay observed with the WT or NES proteins. These findings indicate that nuclear accumulation of E1 is responsible for delaying cell cycle progression and, conversely, that nuclear export of E1 alleviates this effect. Consistent with the need for E1 to accumulate in the nucleus to impair cell cycle progression, we found that an E1 mutant defective for nuclear entry (NLS mutant, Fig. 2B and Table 1) does not interfere with cell cycle progression (Fig. 9A and Table 3).

• Open in new tab
• Download powerpoint

FIG. 9.

Nuclear export of E1 alleviates its detrimental effect on S-phase progression. (A) Cell cycle distribution. C33A cells were transfected with the indicated EYFP-31E1, prepared, and analyzed as described in Fig. 8. The histograms show that nuclear E1, but not cytoplasmic E1, delays S phase. Quantification of these results is reported in Table 3. (B) Colony formation assays (CFA). CFAs were performed with C33A cells transfected with the indicated EYFP-31E1 proteins (expressed from a neomycin-resistant plasmid). After a 3-week selection in G418, the colonies were stained with methylene blue. The results indicate that nuclear but not cytoplasmic E1 impairs colony formation. (C) Cells were transfected with an expression vector for Flag-tagged HPV31 E1 (3F-31E1) and colony formation assayed as described in panel B. An expression vector for Flag-tagged HPV31 E2 (3F-31E2) and the empty vector expressing the three-Flag epitope (3F) alone were used as controls.

To substantiate the cell cycle results presented above, we investigated the impact of E1 on cellular proliferation by using a colony formation assay (CFA). In these experiments, C33A cells were transfected with different EYFP-31E1 expression vectors (neomycin resistance), selected in G418-containing medium for 3 weeks, and the number of drug-resistant colonies was then visualized by methylene blue staining. As can be seen in Fig. 9B, we found that E1 proteins that accumulate in the nucleus (WT, NES, and CBM/NES) drastically reduced colony formation. In contrast, E1 mutants that accumulate predominantly in the cytoplasm (CBM, NLS) had little to no effect (Fig. 9B). Finally, we ruled out any possible effect of the EYFP moiety on cellular progression by confirming the capacity of 3F-31E1, but not of 3F-31E2, to inhibit colony formation in CFA (Fig. 9C).

Altogether, the results presented above indicate that accumulation of E1 in the nucleus, but not in the cytoplasm, impairs cellular proliferation by specifically delaying S-phase progression. Continuous interference with cell cycle progression may therefore explain why the E1 NES mutant protein is unable to maintain the viral episome through many cell divisions (Fig. 7).