Hdm2 negatively regulates telomerase activity by functioning ... - Nature

Oncogene (2010) 29, 4101–4112

& 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc

ORIGINAL ARTICLE

Hdm2 negatively regulates telomerase activity by functioning as an E3 ligase of hTERT W Oh1,4,5, E-W Lee1,4, D Lee2, M-R Yang1, A Ko1, C-H Yoon2, H-W Lee3, Y-S Bae2, CY Choi2 and J Song1 1 Department of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon, Korea; 2Department of Biological Science, Sungkyunkwan University, Suwon, Korea and 3Department of Biochemistry, Yonsei University, Seoul, Korea

In this study, we identified posttranslational regulation of human telomerase reverse-transcriptase (hTERT) by the E3 ligase Hdm2. The telomerase activity generated by exogenous hTERT in U2OS cells was reduced on adriamycin treatment. The overexpressed levels of hTERT were also decreased under the same conditions. These processes were reversed by treatment with a proteasome inhibitor or depletion of Hdm2. Furthermore, intrinsic telomerase activity was increased in HCT116 cells with ablation of Hdm2. Immunoprecipitation analyses showed that hTERT and Hdm2 bound to each other in multiple domains. Ubiquitination analyses showed that Hdm2 could polyubiquitinate hTERT principally at the N-terminus, which was further degraded in a proteasomedependent manner. An hTERT mutant with all five lysine residues at the N-terminus of hTERT that mutated to arginine became resistant to Hdm2-mediated ubiquitination and degradation. In U2OS cells, depletion of Hdm2 or addition of the Hdm2-resistant hTERT mutant strengthened the cellular protective effects against apoptosis. Similar results were obtained with the Hdm2-stable H1299 cell line. These observations indicate that Hdm2 is an E3 ligase of hTERT. Oncogene (2010) 29, 4101–4112; doi:10.1038/onc.2010.160; published online 10 May 2010 Keywords: Hdm2; telomerase; hTERT; p53; ubiquitination

Introduction Telomeres are specialized structures located at the end of eukaryotic chromosomes. They provide cells with genomic stability by protecting chromosomes from fusion, rearrangement and recombination with other chromosomal ends (Maser and DePinho, 2002; Blasco, 2003). Furthermore, a group of protein complexes Correspondence: Professor J Song, Department of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon 440-746, Korea. E-mail: [email protected] 4 These authors contributed equally to this work. 5 Current address: Department of Molecular and Cellular Oncology, University of Texas, MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA. Received 19 June 2009; revised 24 February 2010; accepted 5 April 2010; published online 10 May 2010

including Ku, MRE11, TRF1, TRF2 and POT1 are assembled around telomeres forming D- and T-loops that participate in the regulation of telomere homeostasis (de Lange, 2005). Telomerase is a ribonucleoprotein complex composed of two major components: human telomerase reverse-transcriptase (hTERT) and integral RNA (hTERC or hTR) (Blackburn, 2000). Telomerase is responsible for de novo synthesis of 50 -TT AGGG-30 repeats that are added to the end of telomeres, producing single-stranded 30 -end overhangs (McEachern et al., 2000). The expression of human hTERT is restricted to germ cells, embryonic stem cells and most human cancer cells, even though hTR can be detected in most cells (Blasco, 2003; Harley, 2008). Introduction of hTERT into normal human cells is associated with cellular telomerase activity and induced extension of the life span (Bodnar et al., 1998; Counter et al., 1998; Vaziri and Benchimol, 1998; Luiten et al., 2003; Roth et al., 2003), whereas inhibition of hTERTinduced apoptosis is linked to telomere shortening (Hahn et al., 1999; Zhang et al., 1999; Teng et al., 2003). In addition, hTERT has been implicated in promotion of cell survival and proliferation (Cao et al., 2002). Thus, regulatory pathways of hTERT in transcriptional and posttranslational levels have been considered as a critical factor in tumorigenesis and aging (Urquidi et al., 2000; Cong and Shay, 2008). For example, the transcriptional expression of hTERT is suppressed by p53 and Mad1, whereas Myc and Sp1, Bmi-1, E6 protein and insulin-like growth factor 1 function as positive transcriptional activators (Kyo et al., 2000; Veldman et al., 2001; Akiyama et al., 2002; Dimri et al., 2002). Telomerase itself is also regulated by phosphorylation mediated by Akt, c-Abl and protein kinase C (Cong et al., 2002). MKRN1, an E3 ubiquitin (Ub) ligase, promotes polyubiquitination of hTERT, resulting in suppression of telomerase activity and shortening of telomere length (Kim et al., 2005).

Results Hdm2 induces degradation of hTERT The transcriptional expression of hTERT is suppressed by p53, a potent tumor suppressor (Kusumoto et al., 1999; Kanaya et al., 2000; Xu et al., 2000; Shats et al., 2004).

Hdm2 is an E3 ligase of hTERT W Oh et al

4102

Figure 1 Stress-induced downregulation of hTERT is mediated by Hdm2. (a) Levels of hTERT–HA under stress conditions. U2OS cells were transfected with plasmids expressing hTERT–HA and were treated with 20 mM etoposide and 5 mM adriamycin for 8 h. Cells were treated with 20 mM MG132 for 4 h before harvest. Cells were lysed and immunoblotted with a-HA, a-p53, a-Hdm2 and a-b-actin antibodies. (b) Telomerase activity under stress conditions. U2OS cells were transfected and treated as described in (a). The cell extracts (30 ng) were prepared and were assayed for TRAP analyses (Kim and Wu, 1997). PC denotes positive control. The 36 bp represents the internal TRAP assay standards. TRAP activity was measured by quantifying the relative amounts of bands in the TRAP assays after normalizing to 36 bp band using densitometry (ImageJ Software, NIH, Bethesda, MD, USA). (c) Levels of hTERT–HA under Hdm2 siRNA treatment. U2OS cells were first transfected with control or Hdm2 siRNA for 48 h, followed by transfection of plasmidexpressing hTERT–HA. Adriamycin was treated for 8 h. Lysates were subjected to immunoblotting analyses. (d) TRAP assays following suppression of Hdm2. U2OS cells were transfected and treated as described above. Telomerase activity was analyzed using TRAP assays as described above.

To identify a possible posttranslational regulatory pathway between p53 and hTERT, the latter was transiently transfected into U2OS, a p53-positive osteosarcoma cell line with alternative lengthening of telomere activity. When U2OS cells were treated with chemotherapeutic agents, etoposide or adriamycin, decreased protein levels of transiently expressed hTERT with a concomitant increase in p53 and Hdm2 were observed (Figure 1a, lanes 1–4). The decreased levels of hTERT were reversed by treatment with the proteasome inhibitor MG132 (Figure 1a, lanes 5 and 6). As U2OS expresses TERC, a telomerase RNA component, but no hTERT, the cell extracts appropriately did not show telomerase activity measured by telomeric repeat amplification protocol (TRAP) analyses (Figure 1b, lane 1). In contrast, expression of exogenous hTERT-induced telomerase activity (Figure 1b, lane 2), which was decreased by the addition of the selected chemotherapeutic reagents (Figure 1b, lanes 3 and 4). Telomerase activity was recovered in the presence of MG132 (Figure 1b, lanes 5 and 6). As Hdm2 is a well-known target protein and E3 ligase of p53 (Brooks and Gu, 2006), we suspected that Oncogene

Hdm2 could be involved in the above-noted processes. To ascertain this, the endogenous level of Hdm2 was depleted up to 90% by treating U2OS cells with Hdm2 small interfering RNA (siRNA) under normal or stressed conditions (Figure 1c, lane 2). The levels of overexpressed hTERT were reduced when treated with adriamycin (Figure 1c, lanes 1 and 3). Interestingly, when the levels of endogenous Hdm2 were suppressed in normal or stressed conditions, the levels of hTERT protein were significantly increased (Figure 1c, lanes 2 and 4). TRAP analyses also indicated that the adriamycin-induced suppression of hTERT activity was recovered by more than 2.5-fold by the depletion of Hdm2 (Figure 1d, lanes 3 and 4). We next used the HCT116 human colorectal carcinoma cell line, which harbors endogenous hTERT and p53. Overexpressed hTERT was degraded in a proteasomedependent manner by Hdm2 in HCT116 cells (Figure 2a). We observed the same results in U2OS cells (data not shown). Using Hdm2 siRNA, endogenous Hdm2 proteins were depleted under stressed or normal conditions (Figure 2b). Adriamycin treatment or Hdm2 depletion-induced accumulation of p53 and, at the same

Hdm2 is an E3 ligase of hTERT W Oh et al

4103

Figure 2 Hdm2 induces degradation of hTERT. (a) Hdm2-induced degradation of hTERT. HCT116 cells were treated with plasmids expressing hTERT–HA and Hdm2 (0.3 and 0.9 mg). Cell lysates were analyzed by immunoblotting using the indicated antibodies. Plasmid-expressing green fluorescent protein (GFP) was used for control. (b) Suppression of Hdm2 in HCT116 cells. HCT116 cells were transfected with control or Hdm2 siRNA for 48 h, followed by the treatment with 5 mM adriamycin for 8 h. Cell lysates were analyzed by immunoblotting. For reverse transcription (RT)–PCR analyses, H1299 cells were transfected with control or Hdm2 siRNA as described above. hTERT and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA were detected using RT–PCR. (c) Telomerase activity during depletion of Hdm2. HCT116 cells were transfected with control or Hdm2 siRNA and treated with adriamycin as described above. Cell extracts (30 ng) were prepared and assayed for TRAP analyses (Kim and Wu, 1997). PC denotes positive control. The 36 bp represents the internal TRAP assay standards. TRAP activity was measured by quantifying the relative amounts of bands in the TRAP assays after normalizing to 36 bp band using densitometry (ImageJ software, NIH). (d) Degradation of hTERT by Hdm2 in H1299 cells. H1299 cells were transfected with plasmids expressing hTERT–HA, GFP and Hdm2 (0.2, 0.4 and 0.8 mg) with or without MG132. The steady-state levels of hTERT were detected by western blotting. Plasmid-expressing GFP was used for control. (e) Telomerase activity under Hdm2 expression in H1299 cells. H1299 cells were transfected with increasing amounts of plasmid of Hdm2 (0.2, 0.4 and 0.8 mg). Cell lysates were analyzed for telomerase activity as described above. NC denotes negative control. TRAP analysis for NC was performed in the absence of cell extracts. (f) Effect of adriamycin on telomerase activity in H1299 cells. H1299 cells were treated with or without 5 mM adriamycin (lanes 1 and 2). Cell lysates were analyzed for telomerase activity as above. NC denotes negative control.

time, decreased hTERT mRNA levels (Figure 2b), consistent with previous observations (Rahman et al., 2005). However, the same treatment increased telomerase activity up to twofold under normal conditions and up to fourfold under stress conditions (Figure 2c). These data indicate that hTERT might be regulated posttranslationally. As the endogenous levels of hTERT were not reliably detected using currently available a-hTERT antibodies in HCT116 cells (Wu et al., 2006), we could not investigate further the endogenous levels of hTERT under the aforementioned conditions (data not shown). Finally, we used H1299 human lung cancer cells, which have intrinsic telomerase activity but no p53.

When increasing amounts of Hdm2 were transfected into H1299 cells, exogenous hTERT was degraded in a 26S proteasome-dependent manner (Figure 2d). The intrinsic telomerase activity measured by TRAP was also decreased in an Hdm2 concentration-dependent manner (Figure 2e). On adriamycin treatment, there was no decrease in intrinsic telomerase activity, indicating that telomerase activity suppression did not take place without p53-induced Hdm2 expression (Figure 2f) (Saller et al., 1999). Overall, these results indicate that the stress signals mediated by adriamycin and etoposide can induce Hdm2-mediated degradation of hTERT protein. Accordingly, Hdm2 cells transfected with Oncogene

Hdm2 is an E3 ligase of hTERT W Oh et al

4104

Figure 3 hTERT interacts with Hdm2. (a, b) Interaction between Hdm2 and hTERT. Plasmids expressing Hdm2, hTERT–HA or both were transfected into 293T cells. Whole-cell extracts were prepared and immunoprecipitated with a-HA (a) or a-Hdm2 (b) mouse antibodies and subjected to immunoblotting assays. (c) Immunoprecipitation using a-Hdm2 antibodies, followed by TRAP analyses. Extracts of 293T cells were immunoprecipitated using a-Hdm2 or a-HA antibodies, followed by TRAP assays. Immunoprecipitation of a-HA antibodies was used as a negative control. (d) In vitro binding between hTERT and Mdm2. In vitro-translated Myc–hTERT was subjected to GST pull-down analysis using GST (lane 2) or GST–Mdm2 (lane 3). Bound proteins were detected by western blotting using anti-Myc antibody. CBB denotes GST and GST–Mdm2 used in the GST pull-down assay (lane 4 and 5).

hTERT are able to induce degradation of hTERT in a proteasome-dependent manner in various cell lines. Hdm2 and hTERT have multiple binding sites for each other The observations of Hdm2-induced hTERT degradation led us to test whether Hdm2 could interact with hTERT. Immunoprecipitation analyses using a-HA or a-Hdm2 antibodies verified this interaction (Figures 3a and b). To test whether endogenous proteins were able to bind to each other, we performed immunoprecipitation of endogenous Hdm2 using a-Hdm2 antibodies, followed by TRAP analyses. Immunoprecipitates formed using a-Hdm2 antibodies showed telomerase activities, indirectly indicating the interaction of endogenous Hdm2 and telomerase (Figure 3c). Finally, using recombinant glutathione S-transferase (GST)–Hdm2 and in vitro-translated Myc–hTERT, we identified that both proteins interact with each other (Figure 3d). Domain mapping showed that all three mutants of hTERT were able to bind to Hdm2, suggesting multiple interacting sites on hTERT capable of binding to Hdm2 (Figures 4a and b). To determine the hTERT interaction domain of Hdm2, we constructed four fragments of Hdm2, 1–150, 1–301, 294–491 and 384–491, and assessed their interaction with hTERT. The results indicate that only the 1–301 fragment was able to bind to hTERT (Figures 4c and d). It seems that the Hdm2 region between 150 and 301 contains a motif capable of interacting with hTERT. As Hdm2 fragment 150–301 was not expressed under our condition, we were unable Oncogene

to detect an interaction of this region with hTERT. Collectively, the results support the idea that Hdm2 and hTERT are able to interact with each other, possibly through multiple binding regions. The observation that immunoprecipitates in the presence of a-Hdm2 antibodies showed telomerase activity supports an interaction between telomerase and Hdm2 at endogenous levels. Hdm2 can induce hTERT polyubiquitination Ubiquitination of hTERT was tested using hTERT–HA co-transfected with His–Ub and Hdm2 in the presence or absence of MG132. The samples were pulled down using Ni-NTA-agarose, followed by immunoblotting using a-HA antibodies (Figure 5a). When Hdm2 was coexpressed with hTERT–HA and His–Ub in the presence of MG132, laddered patterns of ubiquitinated hTERT were observed, suggesting that Hdm2-mediated ubiquitination of hTERT could lead to hTERT degradation (Figure 5a, lanes 3 and 4). When mutant Hdm2 D 440–491 (D ring), a mutant lacking E3 ligase activity, was used for ubiquitination analyses, no laddered patterns of polyubiquitinated hTERT were observed (Figure 5b, lanes 3 and 4). This implies that E3 ligase activity of Hdm2 is essential for polyuiquitination of hTERT. For further confirmation, ubiquitination of hTERT was tested using FLAG–hTERT co-transfected with HA–Ub and Hdm2 in the presence or absence of MG132. Immunoprecipitation was performed by aFLAG antibodies and immunoblotting was detected by a-HA antibodies (Supplementary Figure S1a). When

Hdm2 is an E3 ligase of hTERT W Oh et al

4105

Figure 4 hTERT and Hdm2 have multiple binding sites. (a) Schematic diagram of hTERT deletion constructs. (b) Plasmids expressing wild-type (wt) hTERET-HA and its HA-tagged mutants were transfected into 293T cells with plasmid-expressing Hdm2. At 24 h after transfection, the prepared WCE was immunoprecipitated with a-HA mouse antibody and subjected to immunoblotting analyses. (c) The schematic diagrams of Hdm2 deletion mutants. (d) Plasmids expressing Myc-tagged Hdm2 deletion mutants were transfected into 293T cells with plasmid-expressing hTERT–HA. HA immunoprecipitates were eluted by incubating in 2  sample buffer at room temperature for 10 min without boiling. The asterisk indicates immunoglobulins of antibodies.

Hdm2 was coexpressed with FLAG–hTERT and HA– Ub, the amount of ubiquitinated hTERT was increased compared with FLAG–hTERT expressed only with HA–Ub (Supplementary Figure S1a, lanes 2 and 3). In the presence of MG132, a marked increase in ubiquitinated hTERT was evident, suggesting that Hdm2mediated ubiquitination takes place (Supplementary Figure S1a, lanes 5 and 6). Furthermore, HA–Ubtagged proteins were immunoprecipitated, followed by the detection of FLAG–hTERT. The data indicated that hTERT was still ubiquitinated by Hdm2 in the presence of MG132 (Supplementary Figure S1b). When Hdm2 D ring was used, there was no shifted band of polyubiquitinated hTERT, suggesting that Hdm2 E3 ligase activity is required (Supplementary Figure S2a, lanes 3 and 4). The size of Hdm2-mediated ubiquitinated bands of hTERT attained and even exceeded 198 kDa, consistent with polyubiquitination. To confirm this, we used His–Ub7KR, which has all the lysine sites on Ub mutated to arginine, rendering it incapable of being polyubiquitinated (Camus et al., 2007). FLAG–hTERT co-transfected with His–Ub or His–Ub7KR in the presence or absence of Hdm2 was first pulled down with Ni-NTA agarose and immunoblotting was subsequently performed using a-FLAG antibodies to detect hTERT. The results showed clear-laddered patterns of

polyubiquitinated hTERT when hTERT, His–Ub and Hdm2 were expressed together (Figure 5c, lane 3). In contrast, hTERT co-transfected with His–Ub7KR and Hdm2 showed an intensified band around 130 kDa without any laddered bands of larger size (Figure 5b, lane 5). As the sizes of the intensified bands approximately match a monoubiquitinated band of hTERT, it seems that the polyubiquitination process was prevented by the presence of UB7KR. Next, we tested the Hdm2 mutant C464A, which is defective in Hdm2 E3 ligase activity. C464A was able to bind to hTERT (Supplementary Figure S2b), but was unable to induce hTERT degradation (Supplementary Figure S2c). As expected, these mutants were unable to induce Hdm2-mediated ubiquitination of hTERT (Supplementary Figure S2d). Accordingly, the C464A mutant was unable to suppress intrinsic telomerase activity in H1299 cells (Figure 5d). We further observed that Mdm2 could mediate the ubiquitination of hTERT in vitro, whereas the Mdm2 Dring mutant could not (Figure 5e). The data presented in Figures 3c and 5d, and Supplementary Figure S3b show that the interaction between Hdm2 and hTERT was not enough to negatively regulate telomerase activities. To further confirm these, we used Hdm2 fragments 1–301 and 294–491; the former bound to hTERT, whereas the latter did not. Both hTERTOncogene

Hdm2 is an E3 ligase of hTERT W Oh et al

4106

Figure 5 Hdm2 mediates polyubiquitination of hTERT. (a) Ubiquitination of hTERT by Hdm2. H1299 cells were tranfected with plasmids expressing hTERT–HA, His–Ub and Hdm2 in the absence or presence of MG132. His–Ub-conjugated proteins were pulled down using Ni-NTA agarose and detected by western blotting using the indicated antibodies. (b) Ubiquitination of hTERT using Hdm2 E3 ligase-defective deletion mutant. Plasmid-expressing WT Hdm2 and Hdm2D440-491 were co-transfected with hTERT–HA and His–Ub into H1299 cells with MG132. Ubiquitination assay was carried out as in (a). (c) Induction of polyubiquitination of hTERT by Hdm2. Plasmids expressing FLAG–hTERT, His–Ub, His–Ub7KR and Hdm2 were transfected into H1299 cells in the presence of MG132. His–Ub-conjugated proteins were pulled down using Ni-NTA agarose and detected by western blotting. (d) Effect of C464A on telomerase activity. H1299 cells were transfected with plasmid-expressing Hdm2 or C464A, or empty vector. Cell extracts were analyzed for telomerase activity by TRAP assay. (e) In vitro ubiquitination of hTERT by Mdm2. In vitro-translated Myc–hTERT was incubated with a reaction mixture lacking E1, E2 or purified GST–Mdm2, as indicated. The reaction mixtures were resolved by 6% SDS–PAGE and polyubiquitinated hTERT was detected by western blotting using anti-Myc antibody.

binding fragments did not show any effects on hTERT activities (Supplementary Figure S3). Overall, these results were consistent with the suggestion that Hdm2 induces the polyubiquitination of hTERT, and that E3 ligase activity of Hdm2 is essential in these processes. Moreover, the results indicate that a simple binding of Hdm2 fragment or its mutants defective in inducing ubiquitination or degradation of hTERT did not have any effects on hTERT activities. Five lysines on the N-terminus of hTERT are required for Hdm2-mediated ubiquitination As there are multiple binding domains for Hdm2 in the N- and C-termini of hTERT, we next investigated whether Hdm2 could ubiquitinate hTERT deletion fragments. The four N- or C-terminal deletion mutants (Supplementary Figure S4a) were able to bind to Hdm2 (Supplementary Figure S4b). Interestingly, Hdm2 was able to induce polyubiquitination of hTERT deletion Oncogene

mutant 1–542, whereas it did not affect 595–1132 (Figure 6a, lanes 1–6). When 1–350 and 351–1132 fragments were used for further ubiquitination analyses, Hdm2 was only able to induce polyubiquitination on 1–350 (Figure 6a, lanes 7–12). There are five N-terminal lysine sites at residues 78, 94, 236, 339 and 348. All the lysine sites, which are not clustered, as frequently seen for ubiquitination sites, were mutated to arginine residues. When the resulting point mutant, 5KR, was tested for polyubiquitination, Hdm2-mediated ubiquitination was significantly suppressed (Figure 6b). Hdm2 was able to bind to 5KR when FLAG–hTERT was precipitated in the presence of Hdm2 and HA–Ub (Figure 6b, lane 6). It seems that all five sites participate to some degree in ubiquitination, as point mutation of each site individually did not have any major effect on the intensity of Hdm2-mediated ubiquitination (data not shown). When the 5KR mutant was tested for its resistance to Hdm2-mediated degradation in H1299 cells, degradation of 5KR was markedly, but not

Hdm2 is an E3 ligase of hTERT W Oh et al

4107

Figure 6 Five lysines on the N-terminus of hTERT are required for Hdm2-mediated ubiquitination. (a) Ubiquitination of various hTERT deletion mutants by Hdm2. Plasmids expressing HA–Ub or Hdm2 were transfected into H1299 with various plasmids expressing hTERT deletion mutants (1–542, 595–1132, 1–350 and 351–1132). After MG132 treatment, whole-cell extracts were immunoprecipitated with a-FLAG antibodies and immunoblotted with the indicated antibodies. The asterisk in lane 1 of a indicates hTERT fragment 1–542. The lower bands are heavy chains of antibodies. (b) Ubiquitination of hTERT 5KR mutant by Hdm2. Plasmids expressing FLAG–hTERT or FLAG–5KR were co-transfected into H1299 cells with plasmids expressing HA–Ub or Hdm2 in the presence of MG132. WCE was immunoprecipitated with a-FLAG antibodies and subjected to immunoblotting as described above. (c) Degradation analyses of hTERT or 5KR by Hdm2. H1299 cells were transfected with plasmids expressing FLAG–hTERT or FLAG–5KR with or without plasmids for Hdm2. The levels of hTERT were detected by immunoblotting. (d) Effects of Hdm2 on telomerase activity of 5KR. U2OS cells were transfected with plasmids expressing FLAG–hTERT or FLAG–5KR with or without plasmids for Hdm2. Telomerase activities were analyzed by TRAP assays as described in Figure 1b.

completely, suppressed when compared with the degradation of wild-type hTERT (Figure 6c, lanes 2 and 4). These observations suggest that other ubiquitination sites on hTERT might participate in the Hdm2mediated ubiquitination process. To determine whether the observed effects could be correlated with the suppression of telomerase activity, wild-type or 5KR hTERT was transfected into U2OS cells. The transient expression of hTERT in U2OS cells was coincident with telomeric activity, which was suppressed twofold by co-transfected Hdm2 (Figure 6d, lanes 2 and 3, graph). Mutant 5KR, which showed telomerase activity, was not significantly affected by exogenous Hdm2 (Figure 6d, lanes 4 and 5, graph). Collectively, the ubiquitination and degradation data indicate that the five lysines at the N-terminus of hTERT are the major targets for Hdm2-mediated ubiquitination and subsequent degradation. Accordingly, the telomerase activity of 5KR was little affected by coexpressed Hdm2. Anticell death activity of hTERT is suppressed by Hdm2 Extra-telomeric activities of telomerase are associated with cell death (Cong and Shay, 2008). As hTERT

overexpression protects cells from apoptosis (Hahn et al., 1999; Zhang et al., 1999; Teng et al., 2003), we transiently overexpressed FLAG–hTERT in U2OS cells, followed by gating of cells expressing FLAG–hTERT detected by FLAG–fluorescein isothiocyanate antibodies (Supplementary Figure S6a). The exogenous FLAG–hTERT reduced adriamycin-induced cell death from 33 to 14% in U2OS cells when compared with the control (Figure 7b, lower panels 1 and 3). As adriamycin treatment could induce increased levels of Hdm2 (Wang et al., 2009), which would degrade hTERT (Figure 7a, lanes 1–4), we ablated stress-induced Hdm2 proteins by introducing Hdm2 siRNA to the cells (Figure 7a, lanes 5–8). When Hdm2 siRNA treatment was coincident with expression of exogenous hTERT, adriamycininduced cell death was drastically reduced from 42 to 5% (Figure 7b, lower panels 2 and 4). The prevention of cleaved PARP and caspase-3 was consistent with the suggestion that hTERT could more strongly suppress the apoptotic process under Hdm2 depletion (Figure 7a). Next, hTERT–5KR was overexpressed in U2OS cells and tested for adriamycin-induced cell death. On introduction of 5KR, the proportion of Oncogene

Hdm2 is an E3 ligase of hTERT W Oh et al

4108 + - + + -

Hdm2 siRNA

Control siRNA

+ + -

FLAG-hTERT

150 90 60 30

150

2.9%

120

M1

90 60 30

0

33.6%

120 90 60 30

0 200 400 600 800 1000 FL2-A

60

1

2

3

5

4

6

7

M1

8 + + -

hTERT-HA Endo Hdm2

Actin 1

2

3

4

5

6

34.9%

150 120 90 60 30 0

Counts

Cleaved-Caspase-3

Counts

Adriamycin

Cleaved PARP

M1

M1

0 0 200 400 600 800 1000 FL2-A

0 200 400 600 800 1000 FL2-A

3

4

FLAG-5KR

4.2% M1

150 120 90 60 30 0

0 200 400 600 800 1000 FL2-A

0 200 400 600 800 1000 FL2-A 150 120 90 60 30 0

60 30

FLAG-hTERT 150 120 90 60 30 0

3.7% M1

90

R2 gated

pcDNA3 150 120 90 60 30 0

5.1%

120

M1

Counts

+ -

60

13.5% Counts

+ +

Counts

+

Counts

+ -

Untreated

-

0 200 400 600 800 1000 FL2-A 150

13.9%

90

2 R1 gated

Adriamycin hTERT 5KR-HA hTERT-HA

M1

0

0

1

60

0 200 400 600 800 1000 FL2-A

30

0 200 400 600 800 1000 FL2-A

90 30

120

0 0 200 400 600 800 1000 FL2-A

M1

150

41.6%

90 30

0

Actin

60

3.4%

120

0

120

M1

90 30

150 Counts

Cleaved-Caspase-3

150 Counts

Adriamycin

Cleaved PARP

M1

Hdm2 siRNA 150

4.1%

120

0 0 200 400 600 800 1000 FL2-A

Endo Hdm2

Control siRNA 150

5.5%

120

Counts

+

Counts

+ + -

Counts

+ -

Counts

+ +

Counts

+

Untreated

Adriamycin FLAG-hTERT pcDNA3

FLAG-hTERT (R2 gated)

pcDNA3 (R1 gated)

Hdm2 siRNA

Counts

Control siRNA

M1

4.6% M1

0 200 400 600 800 1000 FL2-A 150 120 90 60 30 0

6.2% M1

0 200 400 600 800 1000 FL2-A

0 200 400 600 800 1000 FL2-A

0 200 400 600 800 1000 FL2-A

1

2

3

Figure 7 Depletion of Hdm2 stimulates hTERT-induced cell survival. (a) Expression of hTERT and Hdm2. Transfected U2OS cells were subjected to immunoblotting using the indicated antibodies. (b) Fluorescence-activated cell sorting (FACS) analysis of cell death induced by adriamycin in the absence of Hdm2. U2OS cells were transfected with mock vector or plasmid-expressing FLAG–hTERT with or without depletion of Hdm2 using control or Hdm2 siRNA, followed by treatment with 5 mM adriamycin for 24 h. FLAGnegative (R1) and -positive (R2) cells were first gated as indicated in Supplementary Figure S6 using fluorescein isothiocyanate (FITC)conjugated a-FLAG antibodies and then analyzed for DNA contents by propidium iodide using FACS. The percentage of cells in subG1 is indicated in the figure. (c) Expression of hTERT and Hdm2. Transfected U2OS cells were subjected to immunoblotting as described above. (d) FACS analysis of cell death induced by adriamycin in the presence of hTERT–HA or 5KR–HA. U2OS cells were transfected with mock vector or plasmids expressing hTERT–HA or 5KR–HA. Cells were gated and analyzed using FACS as described above.

cell populations that died was reduced from 35 to 6%, whereas wild-type hTERT reduced population death from 35 to 14% (Figure 7d and Supplementary Figure S6b). In accordance with these data, 5KR was resistant to Hdm2-mediated degradation and prevented cleavage of PARP and caspase-3 (Figure 7c, lanes 5 and 6). To further confirm the role of hTERT and its regulators in suppressing cellular apoptosis, we generated two FLAG–Hdm2 stable H1299 cell lines, SC-7 and SC-10. The tagging of Hdm2 with FLAG did not affect its ability to interact, degrade and ubiquitinate hTERT (Supplementary Figures S5a–c). The stable cell lines constitutively expressed FLAG–Hdm2 and showed suppressed levels of endogenous telomerase activity (Supplementary Figures S5d–e). When the stable control cell line was treated with adriamycin, 50% of the cells were found in the subG1 fraction (Figure 8). In the presence of overexpressed HA–hTERT, approximately 23% of the cell populations died (Figure 8 and Supplementary Figure S6c). When 5KR was overexpressed, the proportion of cell death dropped to 18% (Figure 8). In SC-7 and -10 cell lines, the protective effects of hTERT reduced the rate of cell death from 50 Oncogene

to 33% (Figure 8). Overexpression of 5KR alone was able to further suppress cell death from 50 to 19% (Figure 8). Collectively, these observations indicate that hTERT could have a protective role against apoptosis under the depletion of Hdm2. hTERT 5KR cells not affected by Hdm2 were more efficient than wild type in protecting cells from apoptosis.

Discussion The antagonistic relationship between p53 and hTERT has been generally suggested at the transcriptional level (Kusumoto et al., 1999; Kanaya et al., 2000). In contrast, the posttranslational regulatory link between p53-related mechanisms and hTERT is more tenuous (Cao et al., 2002). In our study, we show that Hdm2, which is a target of p53, is able to function as an E3 ligase of hTERT. The findings that Hdm2 could mediate degradation of hTERT through polyubiquitination reveal how the p53 network could influence telomerase activity on the protein levels. Exogenous

Hdm2 is an E3 ligase of hTERT W Oh et al

4109 5 µM adriamycin

M1 101 102 103 FL2-H

55.2 % M1 101 102 103 FL2-H

Counts

Counts

51.5 %

Counts

101 102 103 FL2-H

HA-5KR 150 120 90 17.9 % 60 M1 30 0 100 101 102 103 104 FL2-H 150 120 90 18.3 % 60 M1 30 0 100 101 102 103 104 FL2-H 150 120 90 19.3 % 60 M1 30 0 100 101 102 103 104 FL2-H

Counts

54.0 % M1

HA-hTERT 150 120 90 23.5 % 60 M1 30 0 104 100 101 102 103 104 FL2-H 150 120 90 32.3 % 60 M1 30 0 104 100 101 102 103 104 FL2-H 150 120 90 34.2 % 60 M1 30 0 104 100 101 102 103 104 FL2-H Counts

Counts

Counts

150 150 120 120 90 90 7.9 % 60 60 M1 30 30 0 0 1 2 3 4 0 10 10 10 10 10 100 FL2-H 150 150 120 120 90 90 7.3 % 60 60 M1 30 30 0 0 100 100 101 102 103 104 FL2-H 150 150 120 120 90 90 8.2 % 60 60 M1 30 30 0 0 1 2 3 4 0 10 10 10 10 10 100 FL2-H Counts

Counts Counts

R2 gated

pcDNA3

HA-5KR

Counts

Counts Counts

Counts

HA-hTERT

150 150 120 120 90 90 8.0 % 7.1 % 60 60 M1 M1 30 30 0 0 100 101 102 103 104 100 101 102 103 104 FL2-H FL2-H 150 150 120 120 90 90 7.6 % 7.4 % 60 60 M1 M1 30 30 0 0 100 101 102 103 104 100 101 102 103 104 FL2-H FL2-H 150 150 120 120 90 90 8.3 % 8.1 % 60 60 M1 M1 30 30 0 0 100 101 102 103 104 100 101 102 103 104 FL2-H FL2-H Counts

H1299/ SC-10

Counts

H1299/ SC-7

Counts

pcDNA3 H1299/ MOCK-2

R1 gated

R2 gated

R1 gated

Counts

Untreated

Figure 8 Expression of hTERT enhances protective effects against cell death. Fluorescence-activated cell sorting (FACS) analysis of H1299 cell lines stably expressing Hdm2. H1299 stable cell lines were transfected with mock vector or plasmids expressing HA–hTERT or HA–5KR, followed by treatment with 5 mM adriamycin for 24 h. HA-negative (R1) and -positive (R2) cells were gated as indicated in Supplementary Figure S6 using fluorescein isothiocyanate (FITC)-conjugated a-HA antibodies and then analyzed for DNA contents by propidium iodide using fluorescence-activated cell sorting (FACS). The percentage of cells in subG1 is indicated.

hTERT introduced into U2OS alternative lengthening of telomere cells could generate telomerase activity, as U2OS cells continually produce hTERC (Cairney and Keith, 2008). This telomerase activity, as well as the protein levels of hTERT, was suppressed by the condition that activated p53 (Figure 1). These processes were inhibited by the presence of proteasome inhibitor or on Hdm2 depletion. Hdm2-mediated ubiquitination and degradation analyses further confirmed that Hdm2 is indeed an E3 ligase of hTERT. Using HCT116 cell lines carrying endogenous p53, we showed that depletion of Hdm2 induced an increase in intact telomerase activity, even with a slight decrease in mRNA levels of hTERT (Figure 2). These results also indirectly suggest that Hdm2 might negatively regulate the levels of hTERT. One interesting finding in this study is that Hdm2 is able to ubiquitinate the five lysine sites at the N-terminus of hTERT. Substitution of these sites with arginine did not disrupt telomerase activity (Figure 6). When 5KR was overexpressed in U2OS cells, followed by adriamycin treatment, it showed a potent cell protective effect against apoptotic stimuli. In contrast, the wild type showed little protective effect (Figure 7). Corroborating these observations, 5KR was much more effective in protecting H1299 cell lines stably overexpressing Hdm2 against adriamycin-induced cell death, when compared with wild-type hTERT (Figure 8). When Hdm2 was depleted in U2OS cells, the protective effects of wild-type hTERT against cell death were also greatly enhanced (Figure 7). Supporting this, adriamycin-mediated hTERT degradation was suppressed in the absence of Hdm2. These results indicate that the protective role of hTERT against cell death can be

annihilated by the presence of Hdm2. The molecular events of anticell death processes accompanying the increased levels of hTERT in most cancer cells are not clearly defined at the moment. However, protective effects of hTERT against apoptotic processes have been described at the molecular level (Kusumoto et al., 1999; Kanaya et al., 2000; Rahman et al., 2005). It seems that a simple binding of Hdm2 onto hTERT cannot induce suppression of telomerase activities. Several experimental results lead to such a conclusion. For example, the Hdm2 E3 ligase-defective mutant Hdm2C464A was able to interact with hTERT but could not induce hTERT ubiquitination or suppression of telomerase activity (Figure 5d, Supplementary Figures S2b–d). The hTERT 5KR mutant resistant to Hdm2 degradation was still able to interact with Hdm2, but its telomerase activities were not suppressed (Figures 6b and d). Furthermore, Hdm2-bound hTERT was functionally active (Figure 3c). Supporting these observations, Hdm2 fragment 1-301, which could bind to hTERT, was unable to induce ubiquitination and degradation, or suppress telomerase activities (Supplementary Figure S3). These results suggest that Hdm2 can regulate hTERT by ubiquitination, followed by degradation, but not merely by binding. The present results have led us to propose a new hypothesis regarding the homeostatic regulation of hTERT. The activation of p53 by apoptotic stimulus generates a large dose of Hdm2, which in turn induces the polyubiquitination of hTERT and its subsequent degradation. In the absence of Hdm2, the cytotoxic effects of p53 are reduced because of the increased levels of hTERT and telomerase activity (Luiten et al., 2003; Rahman et al., 2005). Oncogene

Hdm2 is an E3 ligase of hTERT W Oh et al

4110

Hdm2 interacts with various proteins including Arf, PML, p21, Rb, Hif-1, E2F and p300 (Zhang and Zhang, 2005). A majority of these proteins participate in determining cell survival, as well as death, through an intimate interplay with the p53 network. For example, the levels of Hdm2 increased by p53 lead to ubiquitination-mediated degradation of several antiapoptotic proteins such as IGF-1R, hnRNP K, p21 and Slug (Froment et al., 2008; Bouska and Eischen, 2009; Enge et al., 2009; Wang et al., 2009). Removal of these antiapoptotic factors by Hdm2 induces cells susceptible to the apoptotic process. These observations are similar to Hdm2-mediated degradation of hTERT. As with other substrates of Hdm2, hTERT could also be considered as the most distinguished factor in determining tumorigenesis and aging. The observations that Hdm2 can posttranslationally regulate the activity of telomerase highlight the important and intricate consequences of the p53 regulatory pathway.

Materials and methods Cell culture and transfection Human kidney 293T, colon cancer HCT116, osteosarcoma U2OS and lung carcinoma H1299 cells were maintained in Dulbecco’s modified Eagle’s Medium supplemented with 10% fetal bovine serum (GIBCO-BRL, Grand Island, NY, USA) and 1% penicillin/streptomycin (Invitrogen, Carlsbad, CA, USA) at 37 1C in 5% CO2. HCT116 cell lines were kindly provided by Dr B Vogelstein (Johns Hopkins University, Baltimore, MD, USA). Transient transfections were performed using Lipofectamine 2000 reagent (Invitrogen) or Welfect (Welgene Biotech, Daegu, Korea) according to the manufacturer’s instructions. The siRNA sequence used for the knockdown of Hdm2 was 50 -AATCAGCAGGAATCATCGG AC-30 (Carroll and Ashcroft, 2008). The siRNA sequence used for the knockdown of hTERT was 50 -AACATGCGTCGCAA ACTCTTT-30 (Dairkee et al., 2007). siRNA duplexes were synthesized by Qiagen-Xeragon (Germantown, MD, USA). For control experiments, control siRNA was purchased from Qiagen-Xeragon. siRNA duplexes were transfected using Lipofectamin RNAiMax (Invitrogen) according to the manufacturer’s protocol. U2OS and HCT116 cells were transfected with 30 nM siRNA. At 48 h after transfection, cells were harvested and used for western blot and fluorescence-activated cell sorting analysis. To establish a FLAG–HDM2 stable cell line, pCMV–Tag2B–FLAG–PMLIV was transiently transfected into H1299 cells using Lipofectamine 2000 reagent according to the manufacturer’s instructions (Invitrogen). Multiple independent single clones were selected in 500 mg/ml of G418 and checked for protein expression by immunoblotting or immunofluorescence staining with a-FLAG antibodies. Plasmids pCR3–hTERT–HA and pcDNA3–HA–hTERT fragments (1–350 amino acid (aa), 350–1132aa, 1–542aa, 595–1132aa, 595–946aa and 946–1132aa) have been previously described (Oh et al., 2009). pcDNA3–FLAG–hTERT fragment was generated by polymerase chain reaction (PCR). pcDNA3– FLAG–Hdm2 was PCR amplified and generated by subcloning into the pcDNA3–FLAG vector between EcoRI and XhoI. pcDNA3–FLAG–hTERT 5KR was constructed by the Quickchange site-directed mutagenesis kit (Stratagene, LaJolla, Oncogene

CA, USA) as described by the manufacturer. Wild-type Hdm2 and Dring mutants were kindly provided by J Chen (H Lee Moffitt Cancer Center). Hdm2–C464A and UB7KR were provided by DP Lane (University of Dundee). Myc-tagged Hdm2 fragments were constructed by amplifying fragments using PCR, followed by subcloning into the pCS3–MT–BX vector using EcoRI and XhoI/SalI. Antibodies and chemicals Hdm2 (SMP-14), green fluorescent protein (FL), c-myc (9E10), HA mouse (F-7) and HA rabbit (Y-11) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Hdm2 (Ab-1) from oncogene and hTERT (R1187) were obtained from ACRIS antibodies (Hiddenhausen, Germany). Flag (M2), b-actin, propidium iodide and etoposide were purchased from Sigma-Aldrich (St Louis, MO, USA). Adriamycin was purchased from Duchefa (Haarlem, The Netherlands). Horseradish peroxidase-conjugated antirabbit and anti-mouse antibody were purchased from Stressgen (LaJolla, CA, USA) and Zymed Laboratories (Carlsbad, CA, USA), respectively. MG132 was obtained from Calbiochem (San Diego, CA, USA). Reverse transcription–PCR analysis was carried out as previously described (Oh et al., 2006). The primers used were hTERT mRNA, 50 -GAACT TGCGGAAGACAGTGG-30 (sense), 50 -ATGCGTGAAACC TGTACGCCT-30 (antisense); glyceraldehyde-3-phosphate dehydrogenase, 50 -GAAGGTAAGGTGGGC-30 (sense), 50 -GA AGATGGTGATGGGATTTC-30 (antisense). Immunoprecipitation and western blot analyses Immunoprecipitation and western blot analyses were carried out as previously described (Oh et al., 2006). In brief, cells were lysed in buffer containing 150 mM NaCl, 1 mM EDTA, 0.5 % Triton X-100, 20 mM Tris–HCl (pH 7.5) and protease inhibitor cocktail. Total cell lysates were incubated with specific antibody for 2 h at 4 1C, followed by the addition of protein G-agarose (Amersham Biosciences, Piscataway, NJ, USA). Samples were incubated for 2 h at 4 1C, precipitated, washed three times and dissolved in sample buffer. Total cell lysates and immunoprecipitates were obtained. In vitro pull-down assays In vitro pull-down assays were performed by incubating equal amounts of GST or GST–Mdm2 fusion proteins immobilized onto glutathione-sepharose beads with in vitro-translated hTERT. The mixture was placed on a rocking platform for 90 min in binding buffer (20 mM Tris–HCl, pH 7.5, 1.5 mM MgCl2, 100 mM NaCl, 0.05% Nonidet P-40) and was then washed three times. Bound proteins were eluted and separated by 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), followed by western blotting using anti-Myc antibody. In vivo ubiquitination and deubiquitination assays At 20 h after transfection, 293T cells were treated with 10 mM MG132 for 4 h and lysed. Total cell lysates were incubated with FLAG or c-myc antibodies and then precipitated with G beads at 4 1C for 2 h. Beads containing bound protein were washed and used for the aforementioned SDS–PAGE, nitrocellulose transfer and western blotting. For His pulldown ubiquitination assay, MG132 was added to transfected cells 4 h before harvesting. Cells were washed with phosphatebuffered saline containing 5 mM N-ethyl maleimide and resuspended in denaturing buffer (6 M guanidium chloride, 0.2 M Na2HPO4, 0.2 M NaH2PO4, 1 M Tris, pH 8, 10 mM

Hdm2 is an E3 ligase of hTERT W Oh et al

4111 b-mercaptoethanol, 5 mM N-ethyl maleimide, 5 mM imidazole). Ni-NTA agarose (Qiagen) (75 ml) was added to the resuspended cells and rotated for 16 h at 4 1C. His–Ub-conjugated proteins were sequentially washed four times for 5 min each with buffer A (6 M guanidium chloride, 0.2 M Na2HPO4, 0.2 M NaH2PO4, 1 M Tris, pH 8 and 10 mM b-mercaptoethanol), buffer B (8 M urea, 0.2 M Na2HPO4, 0.2 M NaH2PO4, 1 M Tris, pH 8 and 10 mM b-mercaptoethanol), buffer C (8 M urea, 0.2 M Na2HPO4, 0.2 M NaH2PO4, 1 M Tris, pH 8, 10 mM b-mercaptoethanol and 0.2% Triton X-100) and buffer D (8 M urea, 0.2 M Na2HPO4, 0.2 M NaH2PO4, 1 M Tris, pH 8, 10 mM b-mercaptoethanol and 0.1% Triton X-100). Ni-bound proteins were eluted, heated at 95 1C for 5 min, separated by 6% SDS–PAGE and transferred to a nitrocellulose membrane. Western blotting was performed as described above. In vitro ubiquitination assay His–Ubc5c and GST–Mdm2 fusion proteins were expressed in Escherichia coli BL21(DE3), and were purified by affinity chromatography using His-Bind Quick Column and glutathione-sepharose, respectively. For the in vitro ubiquitination assay, in vitro-translated Myc–hTERT was added to the ubiquitination reaction mixture (final volume 30 ml), which contained 40 mM Tris–HCl, pH 7.6, 5 mM MgCl2, 1 mM adenosine triphosphate, 0.4 mM dithiothreitol, 8 mg of Ub (Sigma-Aldrich), 0.5 mg of E1 (Boston Biochem, Cambridge, MA, USA), 1 mg of His–Ubc5c and 0.2 mg of purified GST– Mdm2. The reactions were incubated for 60 min at 37 1C, and were terminated by boiling in SDS sample buffer. The reaction mixtures were resolved by 6% SDS–PAGE. The ubiquitinated hTERT proteins were detected by western blotting using anti-Myc antibody. Fluorescence-activated cell sorting analysis Cells were harvested using trypsin, fixed with 3.65% formaldehyde and permeabilized with 90% methanol. Cells were washed with phosphate-buffered saline, incubated with fluorescein isothiocyanate-conjugated a-FLAG (F4049, Sigma-Aldrich) or a-HA antibodies (sc7392-fluorescein isothiocyanate, Santa Cruz Biotechnology) for 1 h at 37 1C and stained with 50 mg/ml propidium iodide in the presence of 40 mg/ml RNase A for 1 h at 37 1C. Fluorescein isothiocyanate-

positive cells (FLAG or HA-expressing cells) were first gated as indicated in the Supplementary Figure S6, and then DNA contents were measured using a fluorescence-activated cell sorting Vantage cytofluorometer (Becton-Dickinson, Franklin Lakes, NJ, USA). Data were analyzed using CellQuest Pro software (Becton-Dickinson). TRAP assay TRAP was performed as previously described (Kim and Wu, 1997). Briefly, cells were lysed with 1X CHAPS lysis buffer (Chemicon International, Temicula, CA, USA) containing RNase inhibitor on ice for 30 min. After centrifugation at 13 000 revolutions per minute for 10 min, the protein concentration in the supernatant was determined before incubation at 30 1C for 20 min for telomere extension. The extended products were PCR amplified using TS and ACX primers for 33 cycles (denaturation at 94 1C for 30 s, annealed at 59 1C for 30 s and extended at 72 1C for 60 s). NT and TSNT primers were added as an internal control. The PCR products were separated by 12.5% nondenaturing PAGE in 0.5  TBE. Silver staining was carried out as previously described (Dalla Torre et al., 2002). After electrophoresis, the gels were fixed with a solution containing 0.5% acetic acid and 10% ethanol for 15 min and stained with 0.2% AgNO2 for 10 min. The stained gel was quickly washed and developed with 0.1% formaldehyde and 3% NaOH for 20 min.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements This work was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea (0820110) and by a grant of the Korea Healthcare technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A080333).

References Akiyama M, Hideshima T, Hayashi T, Tai YT, Mitsiades CS, Mitsiades N et al. (2002). Cytokines modulate telomerase activity in a human multiple myeloma cell line. Cancer Res 62: 3876–3882. Blackburn EH. (2000). The end of the (DNA) line. Nat Struct Biol 7: 847–850. Blasco MA. (2003). Telomeres and cancer: a tale with many endings. Curr Opin Genet Dev 13: 70–76. Bodnar AG, Ouellette M, Frolkis M, Holt SE, Chiu CP, Morin GB et al. (1998). Extension of life-span by introduction of telomerase into normal human cells. Science 279: 349–352. Bouska A, Eischen CM. (2009). Murine double minute 2: p53independent roads lead to genome instability or death. Trends Biochem Sci 34: 279–286. Brooks CL, Gu W. (2006). p53 ubiquitination: Mdm2 and beyond. Mol Cell 21: 307–315. Cairney CJ, Keith WN. (2008). Telomerase redefined: integrated regulation of hTR and hTERT for telomere maintenance and telomerase activity. Biochimie 90: 13–23.

Camus S, Menendez S, Cheok CF, Stevenson LF, Lain S, Lane DP. (2007). Ubiquitin-independent degradation of p53 mediated by high-risk human papillomavirus protein E6. Oncogene 26: 4059–4070. Cao Y, Li H, Deb S, Liu JP. (2002). TERT regulates cell survival independent of telomerase enzymatic activity. Oncogene 21: 3130–3138. Carroll VA, Ashcroft M. (2008). Regulation of angiogenic factors by HDM2 in renal cell carcinoma. Cancer Res 68: 545–552. Cong Y, Shay JW. (2008). Actions of human telomerase beyond telomeres. Cell Res 18: 725–732. Cong YS, Wright WE, Shay JW. (2002). Human telomerase and its regulation. Microbiol Mol Biol Rev 66: 407–425. table of contents. Counter CM, Meyerson M, Eaton EN, Ellisen LW, Caddle SD, Haber DA et al. (1998). Telomerase activity is restored in human cells by ectopic expression of hTERT (hEST2), the catalytic subunit of telomerase. Oncogene 16: 1217–1222. Dairkee SH, Nicolau M, Sayeed A, Champion S, Ji Y, Moore DH et al. (2007). Oxidative stress pathways highlighted in tumor cell Oncogene

Hdm2 is an E3 ligase of hTERT W Oh et al

4112 immortalization: association with breast cancer outcome. Oncogene 26: 6269–6279. Dalla Torre CA, Maciel RM, Pinheiro NA, Andrade JA, De Toledo SR, Villa LL et al. (2002). TRAP-silver staining, a highly sensitive assay for measuring telomerase activity in tumor tissue and cell lines. Braz J Med Biol Res 35: 65–68. de Lange T. (2005). Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev 19: 2100–2110. Dimri GP, Martinez JL, Jacobs JJ, Keblusek P, Itahana K, Van Lohuizen M et al. (2002). The Bmi-1 oncogene induces telomerase activity and immortalizes human mammary epithelial cells. Cancer Res 62: 4736–4745. Enge M, Bao W, Hedstrom E, Jackson SP, Moumen A, Selivanova G. (2009). MDM2-dependent downregulation of p21 and hnRNP K provides a switch between apoptosis and growth arrest induced by pharmacologically activated p53. Cancer Cell 15: 171–183. Froment P, Dupont J, Christophe-Marine J. (2008). Mdm2 exerts proapoptotic activities by antagonizing insulin-like growth factor-Imediated survival. Cell Cycle 7: 3098–3103. Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA. (1999). Creation of human tumour cells with defined genetic elements. Nature 400: 464–468. Harley CB. (2008). Telomerase and cancer therapeutics. Nat Rev Cancer 8: 167–179. Kanaya T, Kyo S, Hamada K, Takakura M, Kitagawa Y, Harada H et al. (2000). Adenoviral expression of p53 represses telomerase activity through down-regulation of human telomerase reverse transcriptase transcription. Clin Cancer Res 6: 1239–1247. Kim JH, Park SM, Kang MR, Oh SY, Lee TH, Muller MT et al. (2005). Ubiquitin ligase MKRN1 modulates telomere length homeostasis through a proteolysis of hTERT. Genes Dev 19: 776–781. Kim NW, Wu F. (1997). Advances in quantification and characterization of telomerase activity by the telomeric repeat amplification protocol (TRAP). Nucleic Acids Res 25: 2595–2597. Kusumoto M, Ogawa T, Mizumoto K, Ueno H, Niiyama H, Sato N et al. (1999). Adenovirus-mediated p53 gene transduction inhibits telomerase activity independent of its effects on cell cycle arrest and apoptosis in human pancreatic cancer cells. Clin Cancer Res 5: 2140–2147. Kyo S, Takakura M, Taira T, Kanaya T, Itoh H, Yutsudo M et al. (2000). Sp1 cooperates with c-Myc to activate transcription of the human telomerase reverse transcriptase gene (hTERT). Nucleic Acids Res 28: 669–677. Luiten RM, Pene J, Yssel H, Spits H. (2003). Ectopic hTERT expression extends the life span of human CD4+ helper and regulatory T-cell clones and confers resistance to oxidative stressinduced apoptosis. Blood 101: 4512–4519. Maser RS, DePinho RA. (2002). Connecting chromosomes, crisis, and cancer. Science 297: 565–569.

McEachern MJ, Krauskopf A, Blackburn EH. (2000). Telomeres and their control. Annu Rev Genet 34: 331–358. Oh W, Ghim J, Lee EW, Yang MR, Kim ET, Ahn JH et al. (2009). PML-IV functions as a negative regulator of telomerase by interacting with TERT. J Cell Sci 122: 2613–2622. Oh W, Lee EW, Sung YH, Yang MR, Ghim J, Lee HW et al. (2006). Jab1 induces the cytoplasmic localization and degradation of p53 in coordination with Hdm2. J Biol Chem 281: 17457–17465. Rahman R, Latonen L, Wiman KG. (2005). hTERT antagonizes p53induced apoptosis independently of telomerase activity. Oncogene 24: 1320–1327. Roth A, Yssel H, Pene J, Chavez EA, Schertzer M, Lansdorp PM et al. (2003). Telomerase levels control the lifespan of human T lymphocytes. Blood 102: 849–857. Saller E, Tom E, Brunori M, Otter M, Estreicher A, Mack DH et al. (1999). Increased apoptosis induction by 121F mutant p53. EMBO J 18: 4424–4437. Shats I, Milyavsky M, Tang X, Stambolsky P, Erez N, Brosh R et al. (2004). p53-dependent down-regulation of telomerase is mediated by p21waf1. J Biol Chem 279: 50976–50985. Teng L, Specht MC, Barden CB, Fahey III TJ. (2003). Antisense hTERT inhibits thyroid cancer cell growth. J Clin Endocrinol Metab 88: 1362–1366. Urquidi V, Tarin D, Goodison S. (2000). Role of telomerase in cell senescence and oncogenesis. Annu Rev Med 51: 65–79. Vaziri H, Benchimol S. (1998). Reconstitution of telomerase activity in normal human cells leads to elongation of telomeres and extended replicative life span. Curr Biol 8: 279–282. Veldman T, Horikawa I, Barrett JC, Schlegel R. (2001). Transcriptional activation of the telomerase hTERT gene by human papillomavirus type 16 E6 oncoprotein. J Virol 75: 4467–4472. Wang SP, Wang WL, Chang YL, Wu CT, Chao YC, Kao SH et al. (2009). p53 controls cancer cell invasion by inducing the MDM2mediated degradation of Slug. Nat Cell Biol 11: 694–704. Wu YL, Dudognon C, Nguyen E, Hillion J, Pendino F, Tarkanyi I et al. (2006). Immunodetection of human telomerase reversetranscriptase (hTERT) re-appraised: nucleolin and telomerase cross paths. J Cell Sci 119: 2797–2806. Xu D, Wang Q, Gruber A, Bjorkholm M, Chen Z, Zaid A et al. (2000). Downregulation of telomerase reverse transcriptase mRNA expression by wild type p53 in human tumor cells. Oncogene 19: 5123–5133. Zhang X, Mar V, Zhou W, Harrington L, Robinson MO. (1999). Telomere shortening and apoptosis in telomerase-inhibited human tumor cells. Genes Dev 13: 2388–2399. Zhang Z, Zhang R. (2005). p53-independent activities of MDM2 and their relevance to cancer therapy. Curr Cancer Drug Targets 5: 9–20.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene