Cell-Restricted Immortalization by Human Papillomavirus Correlates with Telomerase Activation and Engagement of the hTERT Promoter by Myc^▿

Plasmids and retroviruses.

We used the following previously described (57, 58) vectors and plasmids: the pJS55 vector, pJS55-16E6, pJS55-16E6-AU1, pJSS55-16E7, the pLXSN vector and pLXSN-16E6, pLXSN-16E7, pLXSN-16E6E7, pGL3-basic (pGL3B), and the pGL3B-hTERT core promoter (previously defined as pGL3B-255). Retrovirus-packaging cells (SD3443 cells) were transfected with a pLXSN vector containing either E6 or E7 or both E6 and E7, as described above, using Lipofectamine 2000 (Invitrogen) as instructed by the manufacturer (54). Culture supernatants containing retroviruses were collected 24 h after transfection.

Cell cultures and generation of stable cell lines.

Primary HFKs and HFFs were cultured from neonatal foreskins as described previously (49) and were maintained in keratinocyte growth media (Invitrogen) supplemented with gentamicin (50 μg/ml) and complete Dulbecco's modified Eagle's medium, respectively. Primary HFKs (passage 2) and HFFs (passage 5) were transduced with amphotropic LXSN retroviruses expressing HPV-16 E6, E7, or both E6 and E7 (see above). Retrovirus-transduced cells were selected in G418 (100 μg/ml) for 5 days. Resistant colonies were pooled and passaged every 3 to 4 days (at a 1:4 ratio for HFKs and a 1:8 ratio for HFFs). HeLa cells were maintained in complete Dulbecco's modified Eagle's medium. All cells were cultured on plastic tissue culture dishes or in flasks.

RT-PCR.

Total cellular RNA was isolated with TRIzol reagent (Invitrogen) and treated with a DNA-free kit (Ambion) according to the manufacturer's instructions. cDNA synthesis and PCR were performed as described previously (28-30). The primers used for reverse transcription-PCR (RT-PCR) were 5′-CAACAAACCGTTGTGTGAT-3′ and 5′-CGTGTTCTTGATGATCTGC-3′ for unspliced E6 (29); 5′-ATGCATGGAGATACACCTAC-3′ and 5′-CATTAACAGGTCTTCCAAAG-3′ for E7; 5′-ATGCCCCTCAACGTTAGCTTC-3′ and 5′-AAGCCGCTCCACATACAGTC-3′ for Myc mRNA; 5′-TGAGCGATAACGATGACATC-3′ and 5′-CATCGAAGGCAGAGATGGTG-3′ for Max; 5′-CGGCGGTTCGGATGAACATC-3′ and 5′-GGTCGATTTGGTGAACGGCT-3′ for Mad1; and 5′-CTCAGACACCATGGGGAAGGTGA-3′ and 5′-ATGATCTTGAGGCTGTTGTCATA-3′ for GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA.

Real-time qRT-PCR.

TaqMan real-time quantitative RT-PCR (qRT-PCR) was performed on a Bio-Rad iCycler MyiQ, using previously reported (10, 29, 30) primers and probes for the quantitation of hTERT mRNA (sense primer, 5′-TGACACCTCACCTCACCCAC-3′; anti-sense primer, 5-CACTGTCTTCCGCAAGTTCAC-3′; and TaqMan probe, 5′-ACCCTGGTCCGAGGTGTCCCTGAG-3′) and Myc mRNA (sense primer, 5′-ACCACCAGCAGCGACTCTGA-3′; anti-sense primer, 5′-TCCAGCAGAAGGTGATCCAGACT-3′; and probe, 5′-GAGGCAGGCTCCTGGCAAAAGGTC-3′). To detect Mad and Max, qRT-PCR was performed, using Bio-Rad Sybr green iQ mixture and the primer described above for RT-PCR. GAPDH was used as an internal control, and expression levels were analyzed using iQ5 software with the normalized expression (ΔΔCT) method according to the manufacturer's (Bio-Rad's) guidelines.

Western blots.

Cells from stable cell lines and cells treated with siRNA duplexes were washed once with phosphate-buffered saline and lysed in 2× sodium dodecyl sulfate (SDS) gel electrophoresis sample buffer. Proteins were separated on a 4% to 20% Tris-glycine gradient gel (Invitrogen) and were electrophoretically transferred to Immobilon-P polyvinylidene difluoride (PVDF) membranes (Millipore). The membranes were blocked in 5% dry milk-phosphate-buffered saline-Tween 20 or 5% bovine serum albumin-phosphate-buffered saline-Tween 20 and incubated with one of the following primary antibodies: anti-p53, anti-Myc, anti-Max, anti-Mad (all from Santa Cruz Biotechnology), pRb (Cell Signaling Technology), or β-actin (Sigma). The secondary antibody was horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG) (Santa Cruz Biotechnology), used at a dilution of 1:10,000. The membranes were visualized by using a Western blotting chemiluminescence luminol reagent (Santa Cruz Biotechnology).

Luciferase assay.

Primary HFKs or HFFs (1 × 10⁵) were seeded onto 24-well plates and grown overnight. Transient transfections were performed, using Lipofectamine 2000 reagent (Invitrogen) according to the protocol provided by the manufacturer. Cotransfections were performed, using 0.5 μg of a core hTERT reporter plasmid (pGL3B-hTERT) and 20 ng of each expression vector (pJS55-16E6, pJS55-16E7, or both), or empty vectors as controls for basal promoter activity. Cells were also cotransfected with 2 ng of plasmid pRL-CMV (Promega), which contains the Renilla reniformis luciferase gene, and used as transfection controls. Firefly and Renilla luciferase activities were measured 24 h after transfection, using the Dual-Luciferase Reporter assay system (Promega).

Real-time q-TRAP.

Human keratinocytes and fibroblasts were lysed and then analyzed by real-time quantitative telomeric repeat amplification protocol (q-TRAP) assays as described previously (10, 29, 30).

ChIP assays.

HFKs, HFFs, or HeLa cells were grown to 80 to 90% confluence in 100-mm dishes. Chromatin immunoprecipitation (ChIP) assays were performed as described previously (58). Normal rabbit IgG (Santa Cruz Biotechnology), rabbit anti-Myc polyclonal antibody (N-262; Santa Cruz Biotechnology), and monoclonal anti-AU1 (Covance) were used for immunoprecipitation assays for HFKs, HFFs, and HeLa cells, respectively. PCR products were separated on a 1.8% agarose gel and visualized by ethidium bromide staining.

E6 and E7 disrupt the p53 and pRb pathways in both HFK and HFF cells.

The HPV-16 E6 and E7 genes are necessary and sufficient to immortalize primary HFKs and ectocervical keratinocytes (20, 38). While E6 directs the ubiquitin-dependent degradation of p53, it also has functions that are p53 independent, including telomerase activation and cell transformation (12, 35, 36). To determine whether the immortalizing activity of E6 and E7 correlated with tumor suppressor inactivation or telomerase induction, we used retroviruses to generate cell strains expressing an empty retroviral vector (pLXSN), E6 (pLXSN-16E6), E7 (pLXSN-16E7), or E6 and E7 (pLXSN-16E6E7), as described in Materials and Methods. The above cell lines were then passed serially in vitro to assay for immortalization. When cells reached 80% confluence, they were split at a 1:4 ratio for HFKs and a 1:8 ratio for HFFs. Therefore, one split would correspond to two or three cell population doublings (PDs) for HFKs and HFFs, respectively.

As shown in Fig. 1, HFKs and HFFs transduced with LXSN ceased proliferating at 22 to 24 PDs and 61 to 64 PDs, respectively. E6 alone extended the life span of HFKs to 32 to 34 PDs and HFFs to 82 to 85 PDs, and E7 alone extended cell divisions to 36 to 38 PDs and 67 to 70 PDs, respectively. Neither E6 nor E7 could independently immortalize HFKs or HFFs. As anticipated, the combined activity of E6 and E7 allowed HFKs, but not HFFs, to bypass senescence and become immortalized. We explored the possibility that the E6 and E7 proteins might be differentially targeting the p53 and pRb pathways in these two different human cell types or differentially inducing the hTERT promoter and thereby activating telomerase.

First, we confirmed that E6 and E7 were expressed in both HFK and HFF cells. RT-PCR was performed with E6- or E7-specific primers. As anticipated, E6 mRNA was expressed in HFKs and HFFs transduced with the E6 and E6E7 retroviruses, while E7 mRNA was expressed in the E7- and E6E7-transduced cells. HFKs and HFFs transduced with the pLXSN vector served as the control, and GAPDH was used to normalize gene expression (Fig. 2A). We also performed reactions without reverse transcriptase to confirm that PCR products came from mRNA, not DNA (data not shown).

To determine whether the E6 and E7 proteins were functional in both cell types, we assayed the levels of tumor suppressors p53 and pRb by Western blotting, using β-actin as a loading control (Fig. 2B). Regardless of cell type, the p53 levels were decreased in E6-transduced cells, and the pRb protein levels were reduced in E7-expressing cells, indicating that the HPV oncoproteins were functional in both cell types and also that inactivation of p53 and pRb is insufficient for immortalization of HFFs.

E6 is sufficient to induce the exogenous hTERT promoter in both HFKs and HFFs.

To determine whether the failure of E6E7 to immortalize HFF cells might be due to the inability of E6 to activate the hTERT promoter, we first analyzed the effect of E6 and E7 on an exogenous hTERT promoter luciferase construct. HFKs and HFFs were transfected with the control vector, pJS55-E6, pJS55-E7, or both, at the same time that they were transfected with the hTERT promoter reporter. In HFKs, E6 induced the hTERT promoter three- to fivefold compared to the control vector (Fig. 3, left side), and E7 increased promoter activity twofold. Together, E6 and E7 enhanced the hTERT promoter activity 7- to 10-fold. Surprisingly, we found that the exogenous hTERT promoter was induced to even greater levels in HFF cells, despite the fact that these cells cannot be immortalized by the E6 and E7 genes (Fig. 3, right side). The level of hTERT promoter activity in the HFF cells was proportionately higher, 9- to 12-, 3- to 5-, and 14- to 18-fold, for E6, E7, and E6E7, respectively.

E6 cannot induce endogenous hTERT transcription or telomerase activity in HFFs.

To validate the studies with the exogenous promoter, we performed qRT-PCR with endogenous hTERT mRNA in HFK and HFF cells. Unexpectedly, we found that neither the E6 nor E6E7 genes could induce the hTERT promoter (Fig. 4A). This finding contrasted with the ability of these genes to effectively induce both the endogenous (Fig. 4A) and exogenous hTERT promoters in HFKs (Fig. 3). To further confirm that this mRNA analysis reflected the downstream activation of telomerase, we prepared lysates from the cells and quantified the telomerase activity by using a q-TRAP assay (see Materials and Methods). As anticipated, E6 increased telomerase activity in HFKs (Fig. 4A). However, consistent with the levels of endogenous hTERT mRNA, we did not detect any increase in telomerase activity in HFF cells. These data indicate not only that E6 and E6E7 are insufficient to induce the endogenous hTERT promoter and telomerase in HFF cells but also that the activity of the transfected hTERT promoter constructs does not reliably reflect the activity of the endogenous promoter when compared between two different cell types.

Myc resides on the endogenous HFK hTERT promoter but not the HFF promoter.

Myc is known to be a direct activator of telomerase in both human keratinocytes and fibroblasts (4, 5, 28, 61), and our previous studies have shown that E6 and Myc physically interact and bind to the hTERT promoter (30, 51, 58). Indeed, the presence of Myc is required for E6-mediated induction of the hTERT promoter. Thus, it was possible that E6 was unable to activate the hTERT promoter because Myc was not present on the hTERT promoter in HFF cells, as it is in noninduced HFK cells.

To test this hypothesis, we performed ChIP assays on both cell types with a Myc antibody. In primary HFKs expressing either the vector or E6, Myc was bound to the endogenous hTERT promoter (Fig. 5A), as reported previously (30, 58). However, we did not observe a signal for Myc binding to the endogenous hTERT promoter in HFFs, even when they expressed E6 (Fig. 5A). Myc was bound to the endogenous hTERT promoter in telomerase-positive HeLa cells (used as a positive control). Therefore, these experiments indicate that Myc is present on the hTERT promoter of telomerase-quiescent HFK cells but not on the promoter of telomerase-quiescent HFF cells. The inaccessibility of Myc to the hTERT promoter in HFF cells could derive from either differences in cellular Myc expression or differences in chromatin structure, which prevent Myc from accessing the promoter. We evaluated Myc expression levels, as discussed in the next section.

Myc protein is expressed at very low levels in HFF cells.

Our initial experiments employed RT-PCR to detect Myc mRNA expression in the two cell types, using both qualitative and qRT-PCR (Fig. 5B and C). In addition, we assayed the abundance of mRNA for two other transcription factors that modulate Myc activity, Max and Mad (11, 16, 21, 31, 39, 44, 46). In Fig. 5B and C, it is apparent that Myc is expressed at the mRNA level in both HFK and HFF cells. It is also apparent that there are similar levels of Mad and Max mRNA in these cells. It is also important to note that E6, E7, and E6E7 do not alter the expression of Myc or the Myc-related proteins.

However, the critical question is whether the Myc protein levels were different between HFK and HFF cells, which we explored by using Western blotting with a Myc antibody. As shown in Fig. 5C, Myc protein was detectable in HFKs expressing vector, E6, E7, or E6E7, and the viral oncoproteins did not alter the expression of Myc or its partners, Max and Mad. Unlike the Myc mRNA studies, we found gross differences in the levels of Myc protein between HFK and HFF cells. That is, there were very low levels of Myc protein in HFF cells compared to that in HFK cells, and it was extremely difficult to detect Myc protein in fibroblasts without the use of MG-132 proteasome inhibitor to prevent Myc degradation. p53 protein was blotted as a control, since it is degraded by E6, and as expected, MG-132 led to an increased level of p53 protein in E6-expressing cells. Thus, our data suggest that the low levels of Myc protein in HFFs might be due to rapid protein turnover. Interestingly, there is a coordinately low level of the Myc-associated Max protein in HFFs, and we could visualize Max reproducibly only when using a proteasome inhibitor. Unlike Myc and Max, however, the Mad protein is expressed at similar levels in both HFK and HFF cells, and there is little or no change in the level of this protein with either E6, E7, or E6E7 or in the presence of a proteasome inhibitor. In summary, the low levels of Myc protein detected in HFF cells might be the etiologic basis for its absence on the endogenous hTERT promoter and the nonresponsiveness of this promoter to E6.

Myc protein expression increases the ability of E6 to engage the endogenous hTERT promoter.

One possibility for the lack of E6's ability to induce the HFF hTERT promoter is that it cannot associate stably with the promoter without Myc binding, which is clearly lacking in HFF cells. To determine if Myc protein might modulate the association of E6 with the hTERT promoter, we transfected both HFF and HFK cells with epitope-tagged E6 (E6-AU1), which is known to retain its ability to induce the hTERT promoter. ChIP experiments with monoclonal anti-AU1 antibody or an IgG control demonstrated that E6-AU1 bound to the hTERT promoter in both types of cells (Fig. 6A), suggesting that E6 might associate with other promoter-associated proteins (e.g., NFX1-91), without the activation of telomerase. However, when the PCR signal was normalized to input, it appeared that E6 binding to the HFK promoter was stronger than that to the HFF promoter (Fig. 6A). This moderate, quantitative difference in E6 binding to the HFK and HFK promoters, however, does not seem to explain the complete absence of hTERT induction by E6 in HFF cells. More likely, it is the absence of Myc on the HFF promoter that is responsible for its lack of responsiveness to E6. Most important, while Myc was sufficient to induce telomerase in both HFKs and HFFs by itself (Fig. 6B), the forced expression of Myc in HFFs significantly increased E6 association with the endogenous hTERT promoter (Fig. 6A, right).

The E6-activated, exogenous hTERT promoter in HFF cells is associated with Myc.

The above data strongly suggest that Myc occupancy on the hTERT promoter correlates with the ability of E6 to induce hTERT transcription. However, as shown in Fig. 3, E6 is sufficient to induce an exogenous hTERT core promoter in HFFs, despite their very low level of Myc protein. This observation provides a unique system to define whether Myc protein levels or the chromatin structure might be the principal determinants of Myc/promoter binding. If Myc is associated with the exogenous hTERT promoter in HFF cells, the “open” chromatin conformation of transfected plasmid DNA may be more accessible to low levels of Myc than the endogenous promoter. If Myc is not associated with the exogenous promoter, it indicates not only that Myc levels regulate hTERT promoter association but also that E6 can induce the hTERT promoter without Myc. To test these possibilities, we transfected the hTERT core promoter and E6 into HFFs and performed a ChIP assay with a Myc antibody. Our data showed that Myc could be detected on the exogenous hTERT core promoter (Fig. 6C), although it was not detectable on the endogenous hTERT promoter (Fig. 5A). Thus, it appears that only hTERT promoters associated with Myc protein are inducible by E6. This conclusion, however, is qualified by the possibility that our ChIP assay conditions may not have been sufficiently sensitive to detect very low levels of Myc on the endogenous promoter.