Telomestatin is a potent G-quadruplex ligand that specifically interacts with the 3′ telomeric overhang, leading to its degradation and that induces a delayed senescence and apoptosis of cancer cells. Protection of Telomere 1 (POT1) was recently identified as a specific single-stranded telomere-binding protein involved in telomere capping and T-loop maintenance. We showed here that a telomestatin treatment inhibits POT1 binding to the telomeric overhang in vitro. The treatment of human EcR293 cells by telomestatin induces a dramatic and rapid delocalization of POT1 from its normal telomere sites but does not affect the telomere localization of the double-stranded telomere-binding protein TRF2. Thus, we propose that G-quadruplex stabilization at telomeric G-overhang inactivates POT1 telomeric function, generating a telomere dysfunction in which chromosome ends are no longer properly protected. (Cancer Res 2006; 66(14): 6908-12)
In human, telomeres consist of the repetition of the G-rich duplex sequence 5′-TTAGGG-3′. A G-rich 3′ strand extends beyond the duplex to form a 130- to 210-base overhang (G-overhang; ref. 1). Telomeres may be structurally organized in different conformations together with several telomere-associated proteins, such as TRF1, TRF2, and POT1 (2). The G-overhang is either accessible for telomerase extension in an open state or inaccessible in a capped (or closed) conformation that involves the formation of a putative t-loop structure (2).
Telomeric proteins stabilize the telomere by protecting the single-stranded G-overhang from degradation (2). Protection of Telomere 1 (POT1) binds specifically to the single-stranded G-overhang (3) and has been described as a regulator of telomere length (4, 5). POT1 has been found associated with the double-stranded telomeric DNA proteins TRF1 and TRF2 through the bridging proteins PTOP/TINT1/PIP1 and TIN2 (6). Suppression of POT1 function by RNA interference in human cells leads to the loss of the telomeric single-stranded overhang, induced apoptosis, senescence, and chromosomal instability in human cells (7, 8).
Because of the repetition of guanines, the G-overhang is prone to form a four-stranded G-quadruplex structure that has been shown to inhibit telomerase activity in vitro (9). Small molecules that stabilize the G-quadruplex are effective as telomerase inhibitors, and several series have been reported to date to induce replicative senescence after long-term exposure to tumor cell cultures (9). Among them, the natural product telomestatin is one of the most active and selective telomeric G-quadruplex ligand (10). It has been recently shown that telomestatin impairs the conformation and the length of the telomeric G-overhang, an effect that is thought to be more relevant than double-stranded telomere erosion as a marker for its cellular activity (11, 12).
Because telomestatin causes cellular effects analogous to those due to dysfunctional telomeric proteins, and because POT1 regulates in vitro the G-quadruplex conformation at telomeres (13), we have investigated the effect of a telomestatin treatment on POT1 binding in vitro and in human cells using a green fluorescent protein (GFP) fusion protein. Our results indicate that G-quadruplex stabilization impairs POT1 binding to telomere sequences and provokes the rapid delocalization of GFP-POT1 from telomeres in human cells.
Materials and Methods
Plasmids. Full-length hPOT1 was cloned into the pET22b expression vector by PCR using the Marathon testis cDNA library (Clontech, Palo Alto, CA). The cDNA was completely sequenced and corresponded to the sequence previously released (3). This construct contained an NH2-terminal T7 sequence, allowing its coupled transcription/transcription. The GFP-POT1 plasmid was constructed by insertion of the POT1 cDNA after PCR amplification from pET22bPOT1 vector at BamHI-XbaI of the pEGFP-C1 plasmid (Clontech).
Electrophoretic mobility gel shift assay. POT1 protein was prepared with the TNT Coupled Transcription/Translation system (Promega, Charbonnières, France) using the pET22bPOT1 vector. Purified TRF2 protein was obtained from an Escherichia coli strain expressing TRF2 fused to a NH2-terminal Tag containing six histidines. Purification of the protein was done by affinity chromatography through a nickel-containing resin (Ni-NTA agarose from Qiagen, Courtaboeuf, France; details on cloning and purification will be published elsewhere). Oligonucleotides were labeled at the 5′ end with T4 polynucleotide kinase (New England Biolabs, Beverly, MA) and [γ-32P]ATP (3,000 Ci/mmol, Amersham Bioscience Europe, Orsay, France) and gel purified.
The binding reactions were done in 20 μL of the following buffer: 50 mmol/L Tris HCl (pH 8), 100 mmol/L KCl, 2 mmol/L MgCl2, 10% glycerol, 0.1% NP40, 20 nmol/L labeled TEL1 probe in the presence of POT1 (2 μg) or TRF2 (50 ng). For POT1 reactions, the SACC1 oligonucleotide (250 nmol/L) and salmon sperm DNA (50 μg/mL) were also added as competitors. For TRF2 reactions, the CXext oligonucleotide (250 nmol/L) that hybridized to the TEL1 overhang was also added. Telomestatin was added at room temperature 15 minutes before the protein. Then, the protein was added, and binding was done for 15 minutes at room temperature. Reaction products were separated by electrophoresis in 6% nondenaturing polyacrylamide gels in 0.25× Tris borate EDTA buffer. The gels were run at 180 V for 1.5 hours, dried on Whatman DE81 paper at 80°C, and visualized by a Phosphorimager (Typhoon 9210, Amersham).
Oligonucleotides used were as follows: TEL1, 5′-TAACCCTAACCCTAAGCGAATTCGTCATGCGAATTCGCTTAGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAGGG-3′; CXext, 5′-GTGCCTTACCCTCTACCCTTACCCTAA-3′; SACC1, 5′-ACTGTCGTACTTGATATGTGGGTGTGTGTGGG-3′; TR2, 5′-TTAGGGTTAGGG-3′.
Cell culture and transfection. EcR293 telomerase-positive cells (14) were grown in DMEM with 100 units of penicillin and 0.1 mg of streptomycin per milliliter and 10% fetal bovine serum (FBS; Invitrogen, Cergy Pontoise, France); 70% to 80% confluence cells were transfected with 5 μg of plasmid in LipofectAMINE 2000 complex in FBS and antibiotic-free DMEM according to the manufacturer (Invitrogen). The medium was replaced after 24 hours, and the cells were grown in DMEM with 100 units of penicillin and 0.1 mg of streptomycin per milliliter containing 400 μg/mL of geneticin. After 15 days of geneticin selection, GFP-positive cells were sorted by fluorescence-activated cell sorting.
Immunofluorescence. For immunofluorescence microscopy, EcR293-GFPPOT1 cells plated on glass coverslips were permeabilized in 0.5% Triton X-100/PBS and fixed with 3% paraformaldehyde. Cells were then washed twice in PBS and treated with permeabilization buffer [20 mmol/L Tris-HCl (pH 8), 50 mmol/L NaCl, 3 mmol/L MgCl2, 300 mmol/L sucrose, and 0.5% Triton X-100] and washed twice with PBS followed by antibody staining with 1 ng/μL for TRF2 4A794 mouse monoclonal (Chemicon-Upstate, Hampshire, United Kingdom) in 0.5% Triton X-100/PBS. The nuclear DNA was stained with 1 μmol/L Hoechst. Secondary antibodies raised against mouse were labeled with Alexa 568 (Molecular Probes, Eugene, OR).
For experiments on living cells, EcR293GFP-POT1 cells were plated on glass coverlips in complete media supplemented with 0.1 μmol/L Hoechst 33342. GFP and Hoechst fluorescence were recorded on a heated stage (37°C) and CO2 chamber of a Axiovert 200 M inverted microscope (Carl Zeiss, Oberkochen, Germany).
Chromatin immunoprecipitations. Chromatin immunoprecipitation was done according to manufacturer procedure (Upstate Biotechnology) using, either TRF2 antibody (H-300, Santa Cruz Biotechnology, Santa Cruz, CA) or with a telomere antibody that recognizes the single- and double-stranded telomere repeats in vitro (19C7, a generous gift from Dr E. Mandine, Sanofi-Aventis, Vitry sur Seine, France). Telomeric sequences in immunoprecipitates were evidenced by PCR amplification according to a previously described method (15). The final telomere primer concentrations were 270 nmol/L (tel1) and 900 nmol/L (tel2), and PCR amplification was subjected to 35 cycles of 95°C for 15 seconds and 54°C for 2 minutes.
Results and Discussion
G-quadruplex formation inhibits POT1 binding to telomeric sequences in vitro. To study the effect of telomestatin on POT1 binding in vitro, a short hairpin oligonucleotide (i.e., TEL1), which reconstitutes the double-stranded telomere with a short 3′overhang, has been used. In the presence of translated POT1, two specific POT1-TEL1 complexes are evidenced by electrophoretic mobility shift assay (EMSA; Fig. 1A, see * and ** bands). These two complexes correspond to the binding of POT1 at the single-stranded extension, as the addition of oligonucleotides (a) CXext that hybridizes to the single-stranded extension and (b) TR2 that titrates POT1 (16) inhibit the formation of these complexes.
In the presence of 2 to 5 μmol/L telomestatin, a strong reduction of the intensity of the signal corresponding to these two complexes is observed, suggesting that telomestatin inhibits the binding of POT1 onto TEL1 (Fig. 1A). Furthermore, the telomestatin treatment induces a dose-dependent inhibitory effect on the formation of the TEL1-POT1 complexes, with an IC50 of 3 μmol/L for the upper complex (*) and an IC50 of 8 μmol/L for the lower complex (**; Fig. 1B). These data suggest that the formation of the complex with the lower mobility is preferentially inhibited by the presence of telomestatin.
The telomestatin treatment also induces in a dose-dependent manner a faster electrophoretic mobility for TEL1 oligonucleotide (Fig. 1A, see TEL1+ at bottom), with a maximal mobility at 10 μmol/L telomestatin, where TEL1+ represents 87% of the total TEL1 oligonucleotide (Fig. 1C).
Interestingly, this effect on TEL1+ correlates with the binding inhibition of POT1 and may result from the interaction of telomestatin with the TEL1 overhang fold in a G-quadruplex structure, in agreement with previous findings (11, 17).
To further establish the selectivity of the telomestatin interaction with telomeric DNA, we also examined its effect on the binding of purified TRF2 to the TEL1 oligonucleotide by EMSA. In the absence of telomestatin, we observe the formation of a specific TRF2-TEL1 complex, as already described (18). In the presence of telomestatin (1.25-10 μmol/L), we did not observe any significant effect of telomestatin (up to 10 μmol/L) on the TRF2-TEL1 complex (Fig. 1D).
These results suggest that telomestatin is able to inhibit in vitro the binding of POT1 to a short telomeric G-overhang but is not able to inhibit the interaction of telomeric proteins, such as TRF2, with the double-stranded telomeric DNA.
GFP-POT1 is localized at telomeres in EcR293 cells. To examine the effects of telomestatin treatment on the binding of POT1 to telomeres in cultured cells, we have designed a GFP-POT1 vector that was transfected in EcR293 cells. Western blot analysis using a GFP antibody indicated that EcR293-GFPPOT1 cells stably express the fusion protein (Fig. 2A). As previously reported, POT1 overexpression in telomerase-positive cells results in telomere length elongation (4). Similar findings are observed in EcR293 cells transfected with either GFP-POT1 or POT1 (Fig. 2B), suggesting that the NH2-terminal fusion with GFP does not alter the functional property of the fusion protein to transduce telomere extension.
POT1 colocalizes at telomeric ends with different telomere-binding proteins (19). To localize GFP-POT1 in EcR293 cells, colocalization experiments have been done on fixed cells by confocal microscopy using a TRF2 antibody. As shown in Fig. 3A, GFP-POT1 colocalizes with almost all the TRF2 dots, suggesting that GFP-POT1 protein is present at telomeres in EcR293 cells. Thus, cells expressing GFP-POT1 fusion protein may be used as models to investigate the cellular effect of telomestatin on POT1 localization.
Treatment with telomestatin delocalizes telomeric GFP-POT1 in EcR293 cells. To investigate the effect of telomestatin on POT1 binding, EcR293 cells expressing GFP-POT1 have been treated for 48 hours with 2 μmol/L telomestatin (Fig. 3A), a concentration and time exposure with the ligand at which most of the cells are still viable, because the IC50 for 2 and 4 days treatment were equal to 7 and 1.8 μmol/L, respectively. Microscopic examination of treated cells indicated a dramatic change in the nuclear organization of GFP-POT1. Telomestatin strongly reduced the GFP-POT1–punctuated signal associated with telomeres compared with untreated controls (compare with TRF2). We also observe the formation of large and diffuse GFP-POT1 aggregates inside the nucleoli (Fig. 3A, see also B). These aggregates are not due to preparation artifacts, as they are also observed in living cells (Fig. 3B). These modifications of the GFP-POT1 localization were detectable within 24 hours and with 1 μmol/L telomestatin, and when the treatment is prolonged for 72 to 96 hours, the GFP-POT1 telomeric foci almost disappeared from nuclei (results not shown, see also living cell experiments in Fig. 3B).
Microscopic examination of living cells confirms these results obtained with fixed cells (Fig. 3B) but also shows that the telomestatin treatment provokes the formation of additional cytoplasmic GFP-POT1 foci that were not or poorly observed in fixed cells. These foci may correspond to proteasome-dependent degradation of GFP-POT1.
All together, our results indicate that telomestatin induces a reduction of the GFP-POT1 telomeric foci in EcR293 cells at noncytotoxic concentrations and in the range to those used to inhibit POT1 binding at telomeric overhang in vitro.
To further establish that POT1 delocalization is specifically due to telomestatin treatment, we also used several anticancer agents (doxorubicin, etoposide, and vinblastin) whose mechanisms of action are not related to G-quadruplex stabilization. At drug concentrations corresponding to the IC50 on EcR293GFP-POT1 cells, we did not observe any effect on the telomere localization of POT1 in viable cells for doxorubicin (10 nmol/L) and etoposide (10 nmol/L; Fig. 3C) and for vinblastin (data not shown). In contrast, cells undergoing apoptosis present a strong decrease of telomeric POT1 staining (data not shown), in agreement with the degradation of telomeric sequences observed during the late stages of apoptosis (20). Therefore, we conclude that POT1 delocalization by telomestatin is an early event occurring before any evidence for apoptosis or cell growth arrest.
Telomestatin does not impair TRF2 binding at telomeres in EcR293 cells. To examine whether the delocalization of POT1 is a consequence of a general effect on the telomere structure, we have determined whether the telomeric localization of TRF2 is also modified by telomestatin. As shown in Fig. 4A (see also Fig. 3A), the telomere localization of TRF2 is not or barely altered by telomestatin treatment in conditions that provokes the GFP-POT1 delocalization in EcR293 cells. The selectivity of the telomestatin effect was also evaluated by chromatin immunoprecipitation experiments using TRF2. In these experiments, the immunoprecipitated telomere sequences were evaluated by specific PCR amplification, as described previously (15). As shown in Fig. 4B, telomestatin treatment of EcR293GFP-POT1 cells to 48 hours and up to 96 hours does not modify the telomere immunoprecipitation by TRF2, in agreement with the immunofluorescence results. Chromatin immunoprecipitation with telomere antibodies were used as internal controls and did not present variation under telomestatin treatment (Fig. 4B).
These results suggest that telomestatin induces a differential effect between POT1 and TRF2 in EcR293 cells and selectively alters POT1 localization, whereas TRF2 is not or poorly affected. Telomestatin was recently shown to completely dissociate TRF2 from telomere in cancer cells but not in normal or immortalized cells (12), a result in agreement with our data on the immortalized EcR293 cell line. Because in vitro experiments using purified TRF2 indicates that telomestatin is inactive to remove TRF2 from double-stranded telomere repeats (Fig. 1D), it is not clear whether the TRF2 dissociation from telomere in cancer cells directly results from a structural alteration induced by telomestatin or is a latter consequence of the telomere destabilization, such as t-loop disruption. Because TRF2 binds to double-stranded telomere and participate to t-loop, a limited effect of telomestatin to telomeric ends would be difficult to detect by microscopy or chromatin immunoprecipitation and is not excluded.
Telomestatin was also reported to induce alterations of the telomeric G-overhang conformation in different tumor cell lines that leads to a degradation of the G-overhang (11, 12). Indeed, telomestatin also provokes a decrease of the G-overhang signal, in EcR293 cells treated for 48 hours at 2 μmol/L (results not shown). This may indicate that the G-overhang alteration preferentially impairs the telomere localization of POT1 in EcR293 cells compared with TRF2.
D. Gomez, T. Wenner, unpublished results.
In conclusion, our results show that the G-quadruplex ligand telomestatin impairs the binding of POT1 to single-stranded telomeric sequences in vitro and alters the nuclear localization of GFP-POT1, suggesting that major modifications of the telomeric end structure are induced by this ligand for which POT1 is a highly sensitive nuclear marker.
Note: D. Gomez is currently at the Institut de Pharmacologie et de Biologie Structurale, Centre National de la Recherche Scientifique UMR 5089, 205 route de Narbonne, 31077 Toulouse, France.
Grant support: Association pour la Recherche contre le Cancer grant 3644 and Ligue Nationale Contre le Cancer (Equipe labellisée 2006).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Drs. E. Gilson, J.L. Mergny, A. Londono-Vallejo, C. Trentesaux, and H. Morjani for helpful discussions.