Edited by Peter M. Howley, Harvard Medical School, Boston, MA (received for review September 30, 2002)
The papillomavirus E6 protein binds and directs the ubiquitin-dependent degradation of the p53 tumor suppressor protein. Independent of this p53-degradative function, however, E6 induces cellular telomerase activity. This increase in enzyme activity reflects E6-enhanced transcription of the human telomerase reverse transcriptase (hTERT) catalytic subunit, but the molecular basis for this transactivation is unknown. In the present study, we demonstrate that E6/Myc interactions regulate hTERT gene expression. Mad protein, a specific antagonist of Myc, repressed E6-mediated transactivation of the hTERT promoter and this repression was relieved by Myc overexpression. The proximal Myc/ Max-binding element (E-box) in the hTERT promoter was the major determinant of both E6 and Myc responsiveness in keratinocytes. E6 did not alter Myc protein expression or Myc/Max association, and the induction of hTERT by Myc/E6 was independent of Myc phosphorylation at Thr-58/Ser-62 within the transactivation domain. However, immunoprecipitation studies demonstrated that endogenous Myc protein coprecipitated with E6 protein and chromatin immunoprecipitation analyses demonstrated that both E6 and Myc proteins bound to a minimal 295-bp hTERT promoter. Only the “high-risk” E6 proteins bound to the hTERT promoter, consistent with their preferential ability to induce telomerase. The observation that E6 associates with Myc complexes and activates a Myc-responsive gene identifies a mechanism by which this oncogene can modulate cell proliferation and differentiation.
Telomerase is a specialized reverse transcriptase that synthesizes repeat DNA sequences at the ends of chromosomes termed telomeres (1). The absence of telomerase activity in most normal human cells results in the progressive shortening of telomeres with each cell division (2–4), eventuating in growth arrest or replicative senescence (3, 5). In contrast to most human somatic cells, immortalized and cancer cells contain detectable telomerase activity and consequently maintain their telomere length and proliferative potential (6–9).
The telomerase enzyme is a ribonucleoprotein complex comprised of two core subunits, a template RNA subunit [human telomerase RNA (hTR)] (10) and a catalytic protein subunit [human telomerase reverse transcriptase (hTERT)] (11, 12), and subunits important for telomere maintenance and stability (1, 13). Although the hTR template subunit is ubiquitously and equivalently expressed in both normal and tumor tissues (14), the hTERT subunit is selectively expressed in a small subset of normal cells (stem cells), tumor tissues, and tumor-derived cell lines (11, 12, 15, 16), indicating that hTERT is the rate-limiting component of telomerase activity. Indeed, ectopic expression of hTERT alone in telomerase-negative cells is sufficient to restore telomerase activity and induce the immortalization of several primary human cell types (17–19).
The E6 oncoprotein of malignancy-associated human papillomavirus type 16 (HPV-16) has recently been shown to activate telomerase activity in epithelial cell types (20, 21), predominantly by inducing transcription of the hTERT gene (22–24). By itself, E6 can immortalize a subpopulation of human mammary epithelial cells (25–27) and, in cooperation with E7, can immortalize primary human foreskin keratinocytes (HFKs) (28, 29). Interestingly, hTERT can substitute for E6 in E6/E7-mediated immortalization of primary HFKs (27), indicating that telomerase activation constitutes a major immortalizing activity of E6. Although E6 protein directs the degradation of the p53 tumor suppressor protein (30, 31) by interactions with E6-associated protein (E6-AP) (32), this p53-degradative function is not required for the induction of cellular telomerase (20) or HFK immortalization (27). Several laboratories, including our own, have shown that E6 increases cellular telomerase activity in primary HFKs by transcriptional activation of the telomerase hTERT gene (22–24) and that this activation can be mediated by a minimal 295-bp hTERT-promoter fragment (22). Interestingly, this core promoter region contains two canonical consensus sequences for Myc-Max binding (-CACGTG-) (33–36) and Myc protein can transactivate the hTERT promoter by direct binding to these E-box elements (37). Mad overexpression represses hTERT transcription (38, 39), presumably by competing with Myc for binding to these same E-box sites.
In the present study, we demonstrate that E6 depends on Myc activity for the induction of hTERT transcription. In transient reporter assays, Mad-protein-repressed E6-induced transactivation of an hTERT promoter and this transcriptional repression was reversed by Myc overexpression. Mutational analysis of the two Myc-binding sites in the 5′ proximal region of the hTERT promoter revealed that both E6 and Myc depend on an intact downstream E-box for high promoter activity. Finally, our study reveals that the cooperation and dependence of E6 on Myc activity most likely derives from their functional interactions at the hTERT promoter.
Cell Culture. HeLa, IMR-90, and A431 cells were maintained in DMEM supplemented with 10% heat-inactivated FBS.
HFKs were cultured from neonatal foreskins as described (40) and maintained in keratinocyte growth media (GIBCO/BRL), supplemented with gentamycin (50 μg/ml). Primary HFKs at passage 8 were infected with amphotropic LXSN retroviruses (Clontech), expressing the HPV-16 E6 and E7 ORFs, separately or together, as described (21). Retrovirus-infected cells were selected in G418 (100 μg/ml) for 5 days, and resistant colonies were pooled.
Plasmids. The WT hTERT reporter plasmid (pGL3B-255), the E-box mutants (pGL3B-Up, pGL3B-Dn, and pGL3B-BM), and the promoterless pGL3B plasmid were kindly provided by I. Horikawa and J. C. Barrett (National Cancer Institute, National Institutes of Health, Bethesda). Double point substitutions were introduced into each E-box element (pGL3B-Up, CACGTG → CGGGTG; pGL3B-Dn, CACGTG → CACCCG), and the sequence of each mutant was confirmed by nucleotide sequencing. The HPV-16 E6 gene and cellular Myc and Mad cDNAs were subcloned into the pJS55 expression vector. To generate the 58A and 62A Myc mutants, the QuikChange Site-Directed Mutagenesis kit (Stratagene) was used according to the manufacturer's instructions and the following oligonucleotide primer sequences were used: 58A, 5′-GAAATTCGAGCTGCTGCCCGCCCCGCCCCTGTCCCCTAGCCGC; 62A, 5′-CTGCTGCCCACCCCGCCCCTGGCCCCTAGCCGCCGCTCCGGGCTC; and 58A62A, 5′-GAAAT-TCGAGCTGCTGCCCGCCCCGCCCCTGGCCCCTAGCCGCCGC. The underline identifies the nucleotide substitution mutation. All mutations were confirmed by sequencing.
Luciferase Assays. For luciferase assays, 4 × 105 telomerase-negative HFK cells were seeded onto six-well plates and grown overnight. Transient transfections were performed by using FuGENE-6 reagent (Roche Molecular Biochemicals) according to the protocol provided by the manufacturer. Cotransfections were performed by using 2 μg of an hTERT reporter plasmid and 2–50 ng of each expression vector as indicated (HPV-16 E6, Myc, Mad, or Myc mutants) or empty vector as control for basal-promoter activity. Firefly luciferase activity was measured 24 h after transfection by using the dual luciferase reporter assay system (Promega). To control for transfection, cells were cotransfected with 10 ng of the pRL-cytomegalovirus plasmid (Promega), which contains the Renilla reniformis luciferase gene under the control of the cytomegalovirus immediate-early enhancer/promoter. Renilla luciferase activity was measured as described above. All luciferase data represent the average of at least three independent experiments.
Immunoprecipitation (IP) and Immunoblotting (IB). For IPs, cells were grown to 80% confluency and lysed in low-stringency lysis buffer (1% Nonidet P-40/PBS; ref. 41) or radioimmunoprecipitation assay buffer containing a mixture of protease and phosphatase inhibitors (200 μM PMSF/5 μg/ml aprotinin/1 μg/ml leupeptin and pepstatin/10 μM NaF and Na3VO4/50 μM sodium pyrophosphate). Subsequent IP reactions with anti-Max Ab (C-124, Santa Cruz Biotechnology), anti-Myc Ab (N-262, Santa Cruz Biotechnology), AU1 mAb (Babco, Richmond, CA), and anti-HPV-16 E6 Ab (N-17, Santa Cruz Biotechnology), and the conditions for gel electrophoresis and poly(vinylidene difluoride) transfer, followed standard protocols (42). IB assays were performed with the above Abs (in addition to phospho-c-Myc-specific Ab, Cell Signaling, Beverly, MA) as described (42).
Nuclear Extraction. Nuclear extracts were prepared from 80% confluent cell cultures as described (43).
Chromatin IP (CHIP) Assays. Primary HFKs were grown to 70–75% confluency and cotransfected with the pGL3–255 hTERT-promoter plasmid or the pGL3B-promoterless plasmid plus HPV-16 E6 and c-Myc expression vectors by using LipofectAmine 2000 (Invitrogen). Cells were fixed with 1% formal-dehyde for 10 min at room temperature, neutralized by addition of glycine (0.125 M for 5 min), washed for 15 min with buffer I (10 mM Tris·HCl, pH 8.0/10 mM EDTA/0.5 mM EGTA/0.25 Triton X-100) and buffer II (10 mM Tris·HCl, pH 8.0/200 mM NaCl/1 mM EDTA/0.5 mM EGTA), and then suspended in 1 ml of lysis buffer (1% SDS/10 mM EDTA/50 mM Tris·HCl, pH 8.0, 1 mM PMSF, 1 μg/ml aprotinin, and 1 μg/ml leupeptin/2 μg of sonicated salmon sperm DNA). The subsequent extraction methods, IP, and PCR amplification protocols have been described (44). For the exogenous hTERT CHIP assay, primers surrounding the minimal hTERT promoter (-255 to + 40) were used with following sequences: 5′-AACAAACTAGCAAAATAGGC-3′ and 5′-AGTACCGGAATGCCAAGCTTAC-3′. For the detection of endogenous hTERT-promoter binding, the following primers located within the minimal hTERT-promoter region were used: 5′-AGTGGATTCGCGGGCACAGA-3′ and 5′-TTCCCACGTGCGCAGCAGGA-3′.
HPV-16 E6 Induction of the Telomerase hTERT Promoter Requires Myc Activity. Recent studies have demonstrated that E6 transactivation of the hTERT promoter does not involve increased Myc protein expression (22, 23). However, to evaluate whether Myc protein might be necessary for E6-mediated transactivation, we coexpressed E6 and Mad expression vectors in telomerasenegative HFKs. The coexpression of Mad was used to inhibit c-Myc activity, because Mad protein is an antagonist of Myc, competing with Myc for the same binding sites in the promoters of Myc-responsive genes (45, 46), including hTERT (39, 47). Transient reporter assays were performed by using a reporter plasmid (pGL3B-255) containing the 295-bp core hTERT-promoter fragment placed upstream of the firefly luciferase gene (22). We observed that E6 expression alone induced hTERT-promoter activity 5-fold higher than the vector control (Fig. 1A). Similar promoter activation was observed with Myc overexpression (Fig. 3C). However, coexpressing Mad protein repressed the ability of E6 to induce high hTERT-promoter activity (Fig. 1 A). The degree of repression correlated with Mad expression levels because increased concentrations of transfected Mad plasmid (2, 10, and 50 ng) led to an increasing reduction of E6-induced promoter activity. To demonstrate that the repressive effects of Mad protein on E6 induction were a specific consequence of inhibiting Myc activity, we coexpressed Myc with the Mad and E6 proteins. Equal plasmid concentrations of cotransfected Mad and E6 resulted in a repression of E6-mediated promoter activation to basal levels. Similarly, Mad repressed Myc-induced promoter activation (data not shown). However, Mad-induced repression on E6 activity was alleviated by coexpressing Myc (Fig. 1B). The Myc plasmid was capable of fully restoring promoter activation by E6. Unexpectedly, high concentrations of the Myc vector resulted in enhanced hTERT-promoter activity that was greater than that observed by E6 expression alone (Fig. 1B). This finding suggests that E6 not only requires the presence of Myc protein, but also cooperates with it to activate the hTERT promoter (also shown in Fig. 3C).
E6-mediated activation of the hTERT promoter is repressed by Mad and restored by Myc. (A) Increasing amounts of Mad plasmid (2, 10, and 50 ng) inhibit E6-induced hTERT-promoter activity. A 295-bp hTERT-promoter fragment (+40 to -255) was cloned into the pGL3-Basic vector (Promega) upstream of the firefly luciferase reporter gene as described (34). The resulting construct (pGL3B-255, see Fig. 2 A) was transfected into keratinocytes with an E6 expression vector, the pRL-cytomegalovirus R. reniformis reporter plasmid (to standardize transfection efficiencies), and increasing amounts of the Mad plasmid. Relative fold activation represents the normalized activity of pGL3B-255 plus empty vector compared to the normalized activity induced by the indicated expression plasmids. (B) Increasing amounts of Myc plasmid (2, 10, and 50 ng) restore E6-induced hTERT-promoter activity. The amount of the E6 and Mad plasmids was kept constant (10 ng of each) in each transfection. Error bars represent the SD of at least three independent experiments.
E6-Mediated Activation of the hTERT Promoter Depends on an Intact Downstream Myc-Binding Site. Previous studies have shown that activation of the hTERT promoter by Myc (35, 37) or E6 (23, 24) depends on two intact E-box elements in the core promoter region of hTERT. To investigate the contribution of individual E-boxes to the induction of this promoter, we performed a series of transient reporter assays in telomerase-negative HFKs. Double point mutations were introduced into the upstream (pGL3B-Up) E-box at -187, the downstream (pGL3B-Dn) E-box at +22, and both E-box sites (pGL3B-BM) simultaneously. These mutations abrogate Myc protein binding (unpublished results). Each promoter construct was transfected into primary HFKs along with the E6 gene (or control vector), and firefly luciferase activity was measured 24 h later. As anticipated, E6 induced a 5-fold increase in the WT hTERT-promoter activity. A similar induction was seen with the upstream E-box mutant (Fig. 2B). In contrast, a 2.5- to 3-fold reduction in promoter activity was observed with the downstream E-box mutant. Simultaneous mutation of both E-boxes did not further reduce promoter activity.
E6 does not require Myc protein phosphorylation at Thr-58/Ser-62 to induce hTERT promoter activity. (A) Domains of the Myc protein and location of two major phosphorylation sites (○). The major phosphorylation sites of Myc, Thr-58, and Ser-62, were replaced with Ala residues by site-directed mutagenesis (•). TAD, transactivation domain; NLS, nuclear localization signal; b, basic domain; HLH, helix–loop–helix domain; LZ, luceine zipper domain. (B) Mutant Myc proteins are phosphorylation-defective. WT and mutant Myc expression proteins were transiently expressed in Cos cells, and Western blot analysis was performed by using a Thr-58/Ser-62 phospho-c-Myc-specific Ab (Cell Signaling). HeLa and HFK served as positive controls and IMR-90 as a negative control. Phospho-Myc blots were stripped and reprobed for total Myc to demonstrate equal expression and loading. (C) Phosphorylation-defective Myc proteins still cooperate with E6 in the induction of the hTERT promoter. Transient reporter assays were performed as in Fig. 2. The hTERT promoter was cotransfected with the E6 and Myc constructs as indicated. Error bars represent the SD of at least three independent experiments.
An intact downstream E-box is required for activation of the hTERT promoter by either E6 or Myc. (A) Schematic representation of the WT and mutant hTERT constructs. The WT hTERT core promoter (pGL3B-255) was mutated at either or both of the E-box elements (Myc-binding sites). Double-point mutations were introduced at each of these sites to generate an upstream mutant (pGL3B-Up), downstream mutant (pGL3B-Dn), or double mutant (pGL3B-BM). The E-box locations are -186 and +22 nt from the transcription start site. Checkered boxes (-CACGTG-) represent the WT E-boxes, black boxes represent the mutant E-boxes, and black bars represent the location of primers. (B) Effect of E-box mutations on E6-induced hTERT-promoter activity. The WT and mutant hTERT promoters were evaluated for transcriptional activity after introduction of the E6 gene (or vector control). Transient reporter assays were performed as described in Fig. 1. (C) Effect of E-box mutations on Myc-induced hTERT-promoter activity. The same protocol as for B except that the Myc gene was transfected with the hTERT-promoter constructs. Error bars represent the SD of at least three independent experiments.
A similar pattern of hTERT-promoter activity was observed when Myc was overexpressed (Fig. 2C). Although mutations in the upstream E-box alone had little or no effect on promoter activity relative to the WT promoter, mutations in the downstream E-box reduced hTERT-promoter activity by 60%. Simultaneous mutations in both E-boxes did not further reduce promoter activity. The downstream E-box site of the hTERT promoter therefore constitutes the major site that is responsive to E6 and Myc in human keratinocytes.
E6-Induced hTERT-Promoter Activity Is Independent of Myc Phosphorylation. There are two major phosphorylation sites in the N-terminal transactivation domain of Myc (Thr-58 and Ser-62) that, in some experimental systems, regulate transcriptional and transforming activities (48, 49). To determine whether phosphorylation of Myc protein at Thr-58 and/or Ser-62 was required for E6 activation of the hTERT promoter, we generated Myc-expression vectors in which the Thr-58 and Ser-62 residues were individually or simultaneously replaced with Ala residues (Fig. 3A). The WT and mutant Myc genes were transfected into COS cells and nuclear extracts were analyzed by IB by using an Ab specific for the Myc phospho-Thr-58/Ser-68 domain or Ab recognizing total Myc protein. WT Myc protein reacted strongly with the phoshospecific Ab whereas the Thr-58, Ser-62, and Thr-58/Ser-62 mutant proteins showed little reactivity (Fig. 3B). The mutations had little or no effect on the total levels of Myc protein.
The functional consequences of the Thr/Ser mutations were evaluated in transient hTERT-promoter assays by cotransfecting HFKs with equal amounts of E6 plus WT Myc or E6 plus each Myc mutant. The expression of E6 or WT Myc alone stimulated hTERT-promoter activity to similar levels, ≈4-fold above vector control (Fig. 3C). As expected from the results of Fig. 1B, the effect of coexpressing E6 and Myc together was additive, increasing promoter activity ≈10-fold. Interestingly, the level of hTERT-promoter activation by E6 plus each Myc mutant was similar to that induced by E6 plus WT Myc. Moreover, neither the single mutants nor the double mutant by themselves had significantly altered transactivational function relative to the WT Myc protein (data not shown), indicating that Myc transcriptional activity is not altered in keratinocytes by N-terminal phosphorylation.
E6 Expression Does Not Enhance Myc–Max Complex Formation. Although the above experiments indicate that Myc phosphorylation is not a factor in E6/Myc cooperativity, it was possible that E6 might alter Myc interactions with its dimerization partner, Max. Myc heterodimerizes with Max to induce transcription of target genes (50, 41). However, Max can also bind to Mad, thereby opposing Myc transcriptional activity by competing for Max binding (45). It was possible that E6 increased Myc transcriptional function by increasing Myc/Max heterocomplex formation rather than Mad/Max complexes. To investigate this possibility, we performed IP/IB assays on exponentially growing E6- and E7-transduced HFKs. To detect cellular Myc/Max complexes, cell lysates (41) were immunoprecipitated with Max Ab and the resultant immunoprecipitates analyzed by IB by using anti-Myc Ab. In three independent experiments, the level of Myc that complexed with c-Max was similar in HFKs expressing or lacking E6 (Fig. 4). Probing the same blots with anti-Max Abs confirmed that equivalent amounts of Max protein were present in the cell lysates (data not shown). IP and IB with a Myc Ab also demonstrated that equivalent amounts of Myc protein were present in the cell lysates, which is consistent with previous findings (22, 23). Thus, E6 did not alter the level of Myc/Max complexes in keratinocytes.
E6 does not alter the levels of Myc–Max heterocomplexes. Proteins were extracted from exponentially growing, transduced HFK cells as described (50) and then IP and IB with either anti-Myc or anti-Max Ab. (Top) Total c-Max expression detected by anti-Max IP and IB. (Bottom) Total c-Myc expression detected by anti-Myc IP and IB. (Middle) The fraction of c-Myc protein complexed with c-Max protein detected by anti-Max IP and anti-Myc IB.
E6 and Myc Proteins Are in a Complex and Both Can Bind the hTERT Promoter in HFK Cells. Although Myc abundance, state of phosphorylation, and Max association do not appear to mediate E6-induced hTERT-promoter activity, it remained a possibility that E6 associated with Myc complexes and modified their function. To evaluate this possibility, IP/IB assays were performed on the same E6- and E7-transduced HFKs shown in Fig. 4. Cellular protein lysates were prepared in radioimmunoprecipitation assay buffer and then IP with Myc, Max, or E6 Ab. IP proteins were separated by SDS/PAGE, and protein blots were reacted with Myc or E6 Ab. As shown in Fig. 5A, E6 Ab was able to coprecipitate Myc protein in cells expressing E6, but not E7, under conditions where normal Myc/Max binding could be observed. Similar results were obtained when E6 and Myc were overexpressed transiently in HFK or COS cells. Fig. 5B shows the coprecipitation of exogenous Myc protein with an epitopetagged, biologically active E6 protein in COS cells, again under conditions where Myc/Max binding can be demonstrated. E6/ Myc association, therefore, can be observed in two different cell types, in stable and transient conditions, and with two different Abs specific for E6. It was not possible to reverse the order of the E6 and Myc Abs in these IP/IB experiments. The etiology of this asymmetry in coprecipitation might reflect altered epitope availability in the complexes or the abundance of the protein being coprecipitated. However, a previous in vitro study has been able to demonstrate the “pull down” of E6 by using a GST-Myc fusion protein (51), suggesting either that appending a larger epitope to Myc is necessary to facilitate E6 detection or that the sensitivity of an in vitro-binding assay is greater than that of our in vivo coimmunoprecipitation studies.
E6 binds to Myc complexes and the hTERT promoter in vivo.(A) Coimmunoprecipitation of E6 and endogenous Myc proteins in stably transduced HFK cells. Protein lysates of E6- and E7-transduced cells were immunoprecipitated with Myc, Max, or E6 Abs, and the precipitates were immunoblotted with E6 or Myc Ab. As anticipated, Myc protein was detectable by IB in both Myc and Max immunoprecipitates. Myc was also detectable in E6-immunoprecipitates (using Abs against E6 protein), but only in those cells expressing the E6 gene and not in those expressing E7. (B) Coimmunoprecipitation of epitope-tagged E6 protein with exogenous Myc in transiently transfected COS cells. COS cells were transfected with the indicated plasmid constructs and analyzed by IP/IB as described in A. The E6 protein was immunoprecipitated with AU1 Ab, which reacts with an appended 6-aa epitope at the E6 C terminus. Myc was detectable in AU1 immunoprecipitates only when E6-AU1 was present. (C) CHIP assay detects Myc and high-risk E6 proteins bound to an exogenous hTERT promoter. (Left) HFK cells were transiently cotransfected with the pGL3B-255 hTERT plasmid, Myc plasmid, and HPV-16 E6 plasmid. After precipitation of protein–DNA complexes with anti-E6 or anti-Myc Ab, PCR amplification was performed by using hTERT-promoter primers (416-bp product). A positive signal was observed with the anti-Myc and anti-E6 Abs, but not with the control rabbit and goat Abs. (Center) Vector-transfected cells failed to give a signal with anti-E6 or anti-Myc Abs. (Right) HFK cells were transfected with the hTERT and Myc plasmids plus either the AU1-tagged HPV-16 or HPV-6b E6 genes. IP was performed with the AU1 Ab. Only the high-risk HPV-16 E6 protein bound the hTERT promoter. Input refers to DNA from transfected HFK cells. (D) CHIP assay detects E6 and Myc proteins bound to the endogenous hTERT promoter. E6 or Myc Ab was used to precipitate endogenous protein–DNA complexes from stable E6- or E7-expressing cells. The complexes were then subjected to PCR amplification to generate a 257-bp product of the hTERT promoter. The anti-E6 Ab generated a positive signal in E6- but not E7-expressing cells. Myc was associated with the hTERT promoter in both E6- and E7-expressing cells.
Although the E6 and Myc proteins are clearly in the same immuno-complex, it was important to investigate whether they were both bound to the hTERT promoter, thereby providing a mechanism by which E6 could directly facilitate hTERT transcription. The hTERT reporter plasmid (or empty vector plasmid) was transfected with the E6 and Myc expression vectors into HFK cells and the association of Myc and E6 with the minimal hTERT promoter was analyzed by CHIP assays. After formal-dehyde cross-linking and sonication, protein-DNA complexes were precipitated with goat anti-E6 or rabbit anti-Myc Abs. The IP chromatin was then subjected to PCR-amplification by using primers specific for sequences surrounding the hTERT-promoter fragment. This primer set produced a 416-bp PCR product from the hTERT plasmid and input DNA, whereas a 137-bp amplified product was produced from the promoterless plasmid (Fig. 5C). As expected from previous studies (37, 47), anti-Myc Ab was able to precipitate the hTERT promoter, indicating the binding of Myc protein to this element. Interestingly, we observed a strong signal in E6 IPs, indicating that E6 can also bind the hTERT promoter (Fig. 5C). No amplified product was detected with either rabbit- or goat-IgG, demonstrating Ab specificity. In addition, no hTERT-promoter sequences were amplified from E6 immunoprecipitates of vector-transfected cells. Thus, these results identify a 295-bp fragment of the hTERT promoter as an E6- and Myc-binding site, the same site that we have shown to be E6- and Myc-responsive.
We also compared the ability of HPV-16 (high risk) and HPV-6b (low risk) E6 proteins to associate to the hTERT promoter. Because we did not have an Ab against the HPV-6b E6 protein, we constructed AU1-tagged E6 proteins for both HPV-6b and HPV-16 and repeated the CHIP assays. Only the high-risk E6 protein bound to the hTERT promoter (Fig. 5C), although Myc protein was associated with the promoter in each case.
The above experiments demonstrated that E6 and Myc bound to a transfected hTERT promoter. To determine whether they could also bind the endogenous hTERT promoter, we performed CHIP assays on keratinocytes stably expressing either E6 or E7, by using sequence-specific primers for the core hTERT promoter. This primer set produced a 257-bp PCR product as demonstrated by the positive-control hTERT plasmid and input DNA lanes (Fig. 5D). We observed a strong hTERT signal (257-bp product) in E6 immunoprecipitates of HFKs transduced with E6 but not E7 (Fig. 5D). Thus, paralleling its ability to activate the keratinocyte hTERT promoter, E6 can associate with the endogenous hTERT promoter. Binding was also observed with the Myc protein, except that Myc was associated with the promoter in both E6- and E7-transduced cells, suggesting that its binding to the promoter is insufficient for inducing transcription.
In this article, we have established a functional role for Myc protein in E6-mediated activation of the telomerase hTERT promoter. In transient reporter assays in HFK cells, Mad reduced E6-mediated activation of the hTERT promoter (Fig. 1 A). Moreover, we demonstrated that this decrease in promoter activity by Mad was the result of a specific repression on endogenous Myc activity because overexpressing Myc restored E6 function to activate the hTERT promoter (Fig. 1B). Of interest, a single, downstream Myc-binding site (E-box element) was necessary for E6-induced activation of the hTERT promoter.
Our findings indicate that E6 and Myc interact functionally to induce hTERT gene expression, and it appears that this cooperativity is caused by their in vivo association in a complex. Most importantly, the Myc and E6 proteins each associate with a minimal, 295-bp promoter fragment of hTERT. The ability of E6 to associate with Myc complexes suggests that it functions to enhance Myc activity, which has already been shown to independently transactivate the hTERT gene. Of interest, although both normal and E7-transduced keratinocytes have Myc bound to the hTERT promoter, they lack telomerase activity and hTERT expression (20–24). Regulation of the hTERT-promoter activity is obviously complex and involves other positive and negative elements.
In an apparent contradictory study, Gross-Mesilaty et al. (51) have shown that a GST-tagged E6 protein can form a complex with Myc, but that this interaction results in increased turnover of Myc, with resultant decreases in protein levels. A possible explanation for this discrepancy is that the apparent degradation of Myc protein by E6 was observed in an overexpression system that does not necessarily reflect the true function of proteins at physiological expression levels. In addition, although our studies were performed with keratinocytes, the natural host cell for HPV infection, Gross-Mesilaty et al. (51) performed their overexpression studies in Cos and NGP cells, a neuroblastoma cell line. This difference in cell type is particularly significant with regard to telomerase activity, because the activation of telomerase by E6 is specific for epithelial cells (20).
The finding that E6 and Myc coassociate suggests several molecular mechanisms for increasing hTERT transcription. It is possible that the binding of E6 to Myc complexes alters Myc conformation, Myc protein modification (other than the phosphorylation that we have monitored), and/or Myc association with positive and negative regulatory proteins. Examination of the hTERT promoter for the presence of the known binding partners for Myc and E6 could help elucidate changes in the components of the transcriptional complex. For example, a recent report indicates that an E6-associated protein, E6-AP, can function independently as a transcriptional coactivator of the nuclear hormone receptor superfamily (52) and the recruitment of this positive regulatory protein to the hTERT promoter by E6 could potentially facilitate activation. Our observation that only the high-risk E6 proteins associate with the hTERT promoter suggests that E6-AP could have a contributory role. Finally, E6 has also been shown to associate with the interferon regulatory factor 3 and p300/CBP transcription factors (53, 54); the recruitment of the latter could potentially augment Myc signaling. The current study defines an in vivo genre of E6 activity that is relevant to the process of cellular immortalization and suggests that other Myc-responsive genes might also be regulated by E6.
We thank Hiroshi Nakai, Elliott Crooke, Frank Suprynowicz, Iruvanti Sunitha, and Gary Disbrow for their advice and suggestions during these studies. We also greatly appreciate the assistance of Vjeko Tomaic with keratinoctye cell cultures. T.V. was supported by National Research Service Award Predoctoral Grant 1F31 CA90203-02, and the project was funded by National Cancer Institute Grant R01 CA53371 (to R.S.).
↵* To whom correspondence should be addressed. E-mail: schleger{at}georgetown.edu.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: hTR, human telomerase RNA; hTERT, human telomerase reverse transcriptase; HPV, human papillomavirus; HFKs, human foreskin keratinocytes; IP, immunoprecipitation; IB, immunoblotting; CHIP, chromatin IP.
