Phosphorylation of TPP1 regulates cell cycle-dependent telomerase recruitment

Abstract

Telomere maintenance is essential for organisms with linear chromosomes and is carried out by telomerase during cell cycle. The precise mechanism by which cell cycle controls telomeric access of telomerase and telomere elongation in mammals remains largely unknown. Previous work has established oligonucleotide/oligosaccharide binding (OB) fold-containing telomeric protein TPP1, formerly known as TINT1, PTOP, and PIP1, as a key factor that regulates telomerase recruitment and activity. However, the role of TPP1 in cell cycle-dependent telomerase recruitment is unclear. Here, we report that human TPP1 is phosphorylated at multiple sites during cell cycle progression and associates with higher telomerase activity at late S/G₂/M. Phosphorylation of Ser111 (S111) within the TPP1 OB fold appears important for cell cycle-dependent telomerase recruitment. Structural analysis indicates that phosphorylated S111 resides in the telomerase-interacting domain within the TPP1 OB fold. Mutations that disrupt S111 phosphorylation led to decreased telomerase activity in the TPP1 complex and telomere shortening. Our findings provide insight into the regulatory pathways and structural basis that control cell cycle-dependent telomerase recruitment and telomere elongation through phosphorylation of TPP1.

In organisms with linear chromosomes, incomplete replication of telomere ends may lead to the loss of genetic information (1, 2). This so-called end replication problem can be further compounded by telomere attrition and cell cycle arrest as a result of nucleolytic degradation and DNA damage repair reactions (3). The telomerase, a ribonucleoprotein with reverse transcriptase activity, synthesizes telomeric DNA repeats to maintain telomere length (4–7), preventing DNA damage responses and cellular senescence (3).

The intimate link between telomere maintenance and DNA replication dictates a close tie between telomerase action and cell cycle. Coupled telomere replication and telomerase-mediated telomere extension in S phase is evident in human cells (8). For example, both the human telomerase RNA (TR) and human telomerase reverse transcriptase (TERT) associate with telomeres during S phase (9, 10). Recruitment of telomerase to telomeres has long been thought to be a rate-limiting step in telomere elongation. However, the mechanisms that regulate telomeric recruitment of human telomerase during specific cell cycle stages remain to be explored.

In addition to the telomerase complex, core telomere proteins telomeric repeat binding factor (TRF) 1, TRF2, TPP1 (formerly known as TINT1, PTOP, or PIP1), the human orthologue of the yeast repressor/activator protein 1 (RAP1), TRF1-interacting nuclear factor 2 (TIN2), and protection of telomeres 1 (POT1) can complex together to associate with telomeres and regulate telomere function (11, 12). Of the six proteins, TPP1 contains an N-terminal oligonucleotide- and oligosaccharide-binding (OB) fold and is tethered to telomeres through interactions with telomere DNA binding proteins TRF1, TRF2, and POT1 (13, 14). Work from us and others has shown that the TPP1 OB fold interacts with the telomerase and is required for telomere recruitment of telomerase and high telomerase processivity (15–19). Recently, key residues that define the interface for human TPP1 OB fold–telomerase interaction have been identified (20–22). However, these experiments were performed in asynchronous cells and did not address how telomerase was recruited to and extend telomeres at late S/G₂. Given that recombinant TPP1 from bacteria is capable of binding telomerase in vitro, additional mechanisms likely exist in vivo for cell cycle-dependent recruitment or regulation of telomerase.

In this report, we demonstrate that multiple sites on TPP1 can be phosphorylated during cell cycle progression. In particular, phosphorylation of Ser111 (S111), a residue located in the telomerase-interacting OB fold, is enriched during S/G₂/M phase of the cell cycle. Blocking S111 phosphorylation by mutating residue S111 led to reduced telomerase association and telomere shortening. Together, these experiments have uncovered a mechanism by which telomerase activity is regulated through protein phosphorylation.

Results

Human TPP1 Is Phosphorylated in a Cell Cycle-Dependent Manner.

Human telomerase has been shown to extend telomeres during S phase (8), suggesting a cell cycle-dependent mechanism for its recruitment. In ciliate, phosphorylation of telomere end binding protein TEBP-β, a homolog of human TPP1, has been shown to regulate cell cycle-dependent telomerase recruitment (23). In budding yeast, regulation of cell cycle-dependent telomerase recruitment by the telomeric protein Cdc13 requires its phosphorylation by cyclin-dependent kinase 1 (Cdk1) (24, 25). Given that human TPP1 interacts with the telomerase and participates in recruiting the telomerase to telomeres, we decided to investigate whether human TPP1 could be phosphorylated during cell cycle progression.

We first synchronized cells expressing Flag-tagged TPP1 by double thymidine block and followed the expression of TPP1 after releasing cells into cell cycle. As shown in Fig. 1 A and B, slower migrating forms of TPP1 became apparent 9 h after release, which corresponded to late S/G₂ phase. These “supershifted” forms of TPP1 were also observed in cells that had been treated with nocodazole to enrich for cells at G₂/M phase (Fig. 1C). Notably, the supershifted TPP1 bands were phosphatase sensitive, indicating that they resulted from phosphorylation (Fig. 1D).

Fig. 1.

Human TPP1 is phosphorylated at late S and G₂/M. (A) HTC75 cells stably expressing Flag-tagged TPP1 (Flag-TPP1) were synchronized by double thymidine block. The migration of TPP1 proteins at different time points following treatment release was analyzed by Western blotting with the indicated antibodies. Asyn, asynchronous cells. Arrows indicate supershifted TPP1 proteins. (B) Flow cytometry analysis of cell cycle stages of cells from A. At 9 h after release, the majority of cells have progressed into late S and G₂/M. In the FACS histograms, F, C, D, and E are user-defined gates. F, cell debris; C, G₀/G₁; D, S phase; E, G₂/M. (C) HTC75 cells expressing Flag-TPP1 were treated with nocodazole (100 ng/mL for 16 h) and analyzed by Western blotting using the indicated antibodies. (D) Flag-TPP1 proteins were affinity purified from nocodazole-treated HTC75 cells and incubated with calf intestine phosphatase before Western blotting. (E) Eight S/TP motifs are highly conserved on TPP1. The phospho-peptide identified by mass spectrometry is also indicated. RD, POT1-recruitment domain. S/T, Ser-rich region. TID, TIN2-interacting domain. (F) Large-scale IP of Flag-TPP1 was performed by using nocodazole-treated HTC75 cells. The immunoprecipitates were then eluted and sequenced by mass spectrometry. (G) Western blot analysis of nocodazole-treated (100 ng/mL for 16 h) HeLa cells expressing various Flag-tagged TPP1 point mutants. 7A, all eight conserved S/T sites but S111 were mutated to Ala. 8A, all eight conserved S/T residues from E were mutated to Ala. 111A, Ser111 to Ala mutation. (H) Analysis of cells from G by using the antibody pS111Ab that recognizes phosphorylated TPP1 S111. The same blot was also probed with the anti-Flag antibody to detect total Flag-TPP1.

Cell cycle-induced protein phosphorylation is often mediated by CDKs, which control some of the most critical cell cycle-dependent events when coupled with various cyclins (26) and prefer substrates with the serine/threonine-proline motif (S/TP) (27). Of the many S/TP sites on human TPP1, eight are highly conserved (Fig. 1E and Fig. S1). To determine which residues on TPP1 were phosphorylated, Flag-TPP1 was purified from asynchronous and nocodazole-treated HTC75 cells and sequenced by mass spectrometry (Fig. 1F). S111 within the OB fold emerged as the major in vivo phosphorylation site identified in our mass spectrometry analysis (Fig. 1E and Figs. S2 and S3). This finding is in agreement with a large-scale quantitative study of mitotic phosphoproteins, where endogenous TPP1 S111 was found to be phosphorylated at G₂/M (28). Human TPP1 has few Lys/Arg residues, and some of the S/TP sites reside in TPP1 tryptic peptides that are too large for mass spectrometry mapping, therefore our experiments may have missed additional phosphorylation sites. Indeed, mutating S111 to Ala had little effect on TPP1 supershift with nocodazole treatment (Fig. 1G), pointing to the existence of additional phosphorylation sites. We then generated TPP1 mutants with single or multiple S/TP site mutations. Significant reduction of TPP1 supershift only became apparent when nearly all eight conserved S/T residues (e.g., TPP1-7A and TPP1-8A) were mutated (Fig. 1G and Fig. S4), supporting the notion that multiple S/TP sites on TPP1 can be phosphorylated during cell cycle progression.

Phosphorylation of TPP1 S111 Is Cell Cycle Regulated.

The important role of TPP1 OB fold in telomerase recruitment suggests functional significance of cell cycle-dependent phosphorylation of S111. To further explore this area, we raised an antibody against phosphorylated S111 (pS111Ab). As shown in Fig. 1H, pS111Ab specifically recognized phospho-S111 on wild-type Flag-TPP1 and TPP1-7A, but not TPP1-S111A, TPP1-8A, or dephosphorylated TPP1. To further study cell cycle-dependent phosphorylation of S111, we synchronized HeLa cells expressing Flag-tagged TPP1-7A with thymidine followed by nocodazole treatment and analyzed S111 phosphorylation by using pS111Ab. As shown in Fig. 2A, phosphorylation of TPP1-7A could be readily detected immediately after release, when the majority of cells were in G₂/M. TPP1 S111 phosphorylation rapidly decreased as the cells transit through G₁. When cells were synchronized with double thymidine block, S111 phosphorylation of wild-type TPP1 could also be detected during S (2–4 h) and late S/G₂ (6–9 h) (Fig. S5). The peak of S111 phosphorylation in S/G₂/M coincides with the time when telomeres become more accessible for replication (29).

Fig. 2.

Phosphorylation on TPP1 S111 is cell cycle regulated. (A) HeLa cells expressing Flag-tagged TPP1-7A were synchronized by using thymidine followed by nocodazole treatment. Lysates from these cells were then used for analysis of the phosphorylation status of S111. Arrow points to expected TPP1 band with phosphorylated S111. Asy, asynchronous cells. Anti-cyclin B1 antibody was used to indicate cell cycle stages. WCE, whole cell extract. (B) HeLa cells expressing Flag-tagged TPP1-7A or TPP1-8A were treated with nocodazole (100 ng/mL) alone or in combination with the indicated kinase inhibitors for 16 h. Lysates from these cells were then analyzed by immunoprecipitation and Western blotting with the indicated antibodies. p38 inhibitor (SB203580; Calbiochem), 20 μM; JNK inhibitor (SP600125; Calbiochem), 50 μM; ERK inhibitor (U0126; Calbiochem), 10 μM; CDK inhibitor (Roscovitine; Cell Signaling Technologies), 40 μM. The anti–phospho-CDK substrate antibody recognizes the phospho-(S/T)PX(R/K) motif.

To further explore the kinase(s) responsible for S111 phosphorylation, we used kinase inhibitors. In addition to CDKs, the S/TP motif is also found in substrates of mitogen-activated protein kinase (MAPK) family (27). The mammalian MAPK family consists of extracellular signal-regulated kinases (ERK), p38, and c-Jun NH₂-terminal kinase (JNK) [also known as stress-activated protein kinase (SAPK)] (30). We therefore selected chemical inhibitors that target major MAPK groups and CDKs and used cells expressing TPP1-7A or TPP1-8A to better visualize TPP1 gel migration changes as a result of changes in S111 phosphorylation. As expected, pS111Ab was able to detect TPP1-7A, but not TPP1-8A, with nocodazole treatment (Fig. 2B, lanes 3 and 4). Although p38 and ERK inhibitors (16 h treatment) had little effect on S111 phosphorylation (Fig. 2B, lanes 5 and 7 and Fig. S6B), the addition of JNK and/or CDK inhibitors (16 h treatment) to cells abolished S111 phosphorylation (Fig. 2B, lanes 6 and 8). This loss of phosphorylation is unlikely due to changes in cell cycle profile, because similar results were obtained when cells were treated with inhibitors for shorter amount of time (2-4 h) (Fig. S6A). The JNK inhibitor SP600215 has been shown to act on JNK as well as CDKs (31). Indeed, treatment with SP600215 resulted in decreased phosphorylation of CDK substrates as well, as shown with a pan-CDK substrate antibody (Fig. 2B, lane 6). These observations indicate that S111 is likely phosphorylated by CDKs.

Phosphorylation of TPP1 S111 Is Implicated in Telomere Length Regulation.

TPP1 is a critical component of the core telomere-associating complex, and its dysfunction can lead to uncapped telomere ends and disrupted telomere length regulation (15). To probe the role of S111 phosphorylation in telomere maintenance, we generated cells expressing TPP1-S111A, or TPP1 phospho-mimic mutants TPP1-S111D and TPP1-S111E. Wild-type and mutant TPP1 proteins were expressed at comparable levels and retained their ability to interact with TIN2 and POT1 (Fig. 3A), indicating that changes in S111 phosphorylation status did not disrupt telomeric protein complex formation. This result is consistent with the fact that S111 resides outside of both TIN2 and POT1-binding domains (RD and TID domains) on TPP1 (Fig. 1E). Although exogenous TPP1 proteins were expressed at levels higher than endogenous TPP1, cells expressing TPP1 S111 mutants did not exhibit changes in TIF formation (Fig. S7), suggesting that TPP1-mediated telomere end protection does not involve S111 phosphorylation. However, mutations of S111 on TPP1 led to altered telomere length (Fig. 3 B and C). Here, ectopic expression of TPP1-S111A resulted in telomere shortening (∼5 kb to ∼3.5 kb), whereas S111E substitution accelerated telomere elongation (∼5 kb to ∼6.5 kb). The TPP1 S111D mutant did not appear to affect telomere length. It is possible that S to E substitution more closely mimics S111 phosphorylation than S to D substitution. These data suggest that S111 phosphorylation is important for telomere length control in vivo.

Fig. 3.

Phosphorylation of TPP1 S111 is important for regulation of telomere length and TPP1-associated telomerase activity. (A) HTC75 cells stably expressing Flag-tagged wild-type or S111 phosphorylation mutants of TPP1 were analyzed by Western blotting (Upper) or coimmunoprecipitation with anti-Flag antibody. (B) Quantification of average telomere length of cells from A. (C) Representative gel of TRF analysis. PD, population doubling. (D) Synchronized HeLa cells (thymidine followed by nocodazole) expressing Flag-tagged wild-type TPP1 were analyzed for DNA content (Left), and for TPP1-associated endogenous telomerase activity during different cell cycle stages by immunoprecipitation and Q-TRAP (Right). Results from Q-TRAP were normalized based on TPP1 protein amount. Error bars indicate SEs (n = 3). Asy, asynchronous cells. (E) Nocodazole-treated (100 ng/mL for 16 h) HeLa cells expressing wild-type or phosphorylation mutants of TPP1 were similarly analyzed as in D. Results from Q-TRAP were normalized based on protein amount from Western blotting. Error bars indicate SEs (n = 3).

Phosphorylation of TPP1 S111 Regulates Telomerase Activity.

Given that S111 lies within the OB fold that mediates TPP1 interaction with telomerase and that phosphorylation of S111 is cell cycle dependent, we next examined whether the association of TPP1 with telomerase varied throughout cell cycle. HeLa cells expressing Flag-TPP1 were synchronized with thymidine followed by nocodazole treatment, and the level of telomerase activity that associated with TPP1 was assayed by real-time quantitative PCR after release of cells into cell cycle. As shown in Fig. 3D, when cells were in late S and G₂/M, TPP1 brought down a high level of telomerase activity. Notably, this pattern coincides with that of S111 phosphorylation during cell cycle progression (Fig. 2A), raising the possibility that S111 phosphorylation may directly regulate telomerase activities that associates with TPP1.

To further study the effect of S111 phosphorylation on TPP1-associated telomerase activity, we compared the ability of wild-type TPP1 and various TPP1 mutants to bind endogenous telomerase. We took advantage of the fact that changes in TPP1 phosphorylation did not affect its telomere targeting or abilities to interact with TIN2 and POT1 (Fig. S8) and went on to immunoprecipitate Flag-TPP1 and TPP1 mutants from human cells and assay for telomerase activities in the immunoprecipitates. As shown in Fig. 3E, in asynchronous cells, wild-type TPP1 and TPP1-S111A pulled down similar amount of endogenous telomerase activity. In contrast, TPP1-S111A consistently brought down less telomerase activity than wild-type TPP1 when treated with nocodazole (Fig. 3E), suggesting that TPP1-S111A is partially defective in recruiting higher telomerase activity with nocodazole treatment. Similar to wild-type TPP1, the TPP1-7A mutant could still associate with telomerase activity, with the highest level at G₂/M. Both wild-type TPP1 and TPP1-7A were able to pull down higher levels of telomerase activity in response to nocodazole treatment compared with untreated samples (Fig. 3E, Right). In contrast, we failed to detect similar increase in TPP1-associated telomerase activity with the TPP1-8A mutant in the presence of nocodazole. Collectively, these results suggest that the increased telomerase activity at S/G₂/M, rather than basal telomerase activity, in the TPP1 complex depends on S111 phosphorylation.

S111 and Additional Conserved Residues in Human TPP1 OB Fold Form the Interface Important for TERT–TPP1 Interaction.

It has been reported that the TPP1–POT1 complex can recruit and stimulate telomerase activity through direct interaction between TPP1 and telomerase. This interaction depends on the OB fold of TPP1 (15–18, 20). Recent studies have identified the residues that are required for the interface of TPP1–telomerase interaction (20–22). Our results thus far support the model that phosphorylation of S111 increases the telomerase activity that complexes with TPP1. We reasoned that this S111-dependent enhancement became possible because of the structural proximity of S111 to the TPP1–telomerase interaction interface. To test this hypothesis, we performed structural analysis of TPP1 based on the published crystal structure of human TPP1 OB fold (17) (Fig. 4A). Consistent with previous reports (20–22), several of the residues that are required for the interface of TPP1–telomerase interaction, including D166, E168, E169, E171, and E215, reside on the same face of the TPP1 3D structure (Fig. 4A). If mutations of these residues result in the loss of TPP1–telomerase interaction, these mutations should also diminish the amount of telomerase activities associated with TPP1 in our assays. To determine this possibility, we generated Ala mutants and compared their abilities to bring down telomerase activities (Fig. 4 B and C). Because Est3, a yeast telomerase holoenzyme component, interacts with the telomerase through a domain predicted to be similar to TPP1 OB fold (32, 33), we mutated two residues found to be conserved among yeast Est3 proteins and TPP1 (W98 and D148) and used these TPP1 mutants as controls. Similar to mutations of the corresponding residues in yeast Est3 proteins, mutations of W98 and D148 reduced telomerase activities that associated with TPP1. Our results also showed that Ala mutation of surface residues D166/E168, E169/E171, or E215 significantly reduced the level of telomerase activity that associated with TPP1 (Fig. 4 B and C). Importantly, the telomerase activity associated with TPP1-D166A/E168A was lower than wild-type TPP1 throughout the cell cycle, which is consistent with the role of D166 and E168 in regulating TPP1–telomrase interaction (Fig. S9). Telomerase activity that associated with TPP1-D166A/E168A appeared insensitive to nocodazole treatment, suggesting that D166/E168-dependent TPP1–telomerase interaction is a prerequisite for S111 phosphorylation-mediated regulation. Interestingly, when phosphorylated S111 was modeled in the TPP1 OB crystal structure, the phosphate group appeared positioned at the same telomerase interaction interface formed by D166, E168, E169, E171, and E215 (Fig. 4A). This result suggests that phosphorylated S111, along with these D/E residues, participates in TPP1–telomerase interaction and regulates the amount of telomerase activity associated with TPP1.

Fig. 4.

Conserved residues in TPP1 OB fold are important for TPP1–telomerase interaction. (A) Modeling of phosphorylated S111 on crystal structure of the hTPP1 OB fold (PDB ID code 2146) illustrated by the program Pymol. The side (Left) and bottom views (Right) of the human TPP1 structure (represented as the surface and ribbon model with residues 90–97 removed) showing the critical residues predicted to be involved in TPP1 interaction with telomerase. The six residues, phosphorylated S111, D166, E168, E169, E171, and E215 (side chains represented as sticks) are located on the same side of the OB fold and may form the contact surface for the interaction. (B and C) HeLa cells expressing various TPP1 point mutants were analyzed by Q-TRAP as described in Fig. 3. The OB-fold deletion mutant TPP1-∆OB was included as a negative control. Error bars indicate SEs (n = 3). P values were determined by the Student t test. (D) A model of cell cycle-dependent telomerase recruitment and regulation.

Discussion

One of the known mechanisms for modulating telomerase activity is through regulated intracellular trafficking of the telomerase enzyme (9, 10, 34). Although both human TERT and human telomerase RNA have been shown to associate with telomeres only in specific cell cycle stages, the mechanisms of such recruitment remain unclear. It has been demonstrated that TPP1 can function to specifically recruit the telomerase when complexed with POT1 on the single-stranded 3′ region of telomeres (15, 17, 19). Our study here has revealed a mechanism by which TPP1 controls cell cycle-dependent telomerase recruitment.

We found that residue S111 on TPP1 is phosphorylated during S to G₂/M phases of the cell cycle, which coincides with the time window when telomeres become accessible (29). Phosphorylation of S111 on TPP1 appears critical for cell cycle-dependent association between TPP1 and telomerase. First, TPP1 S111A was able to pull down less telomerase activity than wild-type TPP1 in G₂/M. In addition, both wild-type TPP1 and TPP1-7A (in which S111 remains intact) could associate with the highest amount of telomerase activity immediately following nocodazole treatment (G₂/M phase) (Fig. 3 D and E), consistent with S111 phosphorylation pattern (Fig. 2A). Moreover, nocodazole treatment enhanced TPP1-7A, but not TPP1-8A, in its ability to associate with telomerase activity. Uncoupling this cell cycle-dependent event by blocking S111 phosphorylation resulted in telomere shortening, underlining the functional importance of this phosphorylation event. We also found that TPP1 was phosphorylated at multiple S/TP sites and such phosphorylation affected TPP1 mobility, and we cannot rule out the possibility that phosphorylated residues other than S111 may also contribute to telomerase regulation. Given the sharp decrease in S111 phosphorylation as cells entered G₁ phase, it is possible that protein phosphatases may play a critical role in TPP1 regulation as well.

It has been shown in several organisms that telomere elongation coincides with semiconservative telomere replication in late S/G₂ phase (35, 36). For example, telomerase recruitment in ciliate is regulated by phosphorylation of the TPP1 homolog TEBP-β during S phase (23). In Saccharomyces cerevisiae, Cdc13 interacts with the telomerase complex and promotes its telomeric recruitment in a cell cycle-dependent manner (37, 38). Here, Cdk1-dependent phosphorylation of Cdc13 at Thr308 appears essential for efficient telomerase complex recruitment to telomeres in late S/G₂ and telomere length maintenance (24). Compared with yeast, we present evidence here that human cells use TPP1 for regulating cell cycle-dependent telomerase activity, and it is interesting to note that the predicted structure of yeast Est3 is similar to the crystal structure of the TPP1 OB fold (32, 33). Given that CDK inhibitors blocked TPP1 S111 phosphorylation, CDKs may be the kinases responsible for phosphorylating S111 in S or G₂/M. Recently, human conserved telomere capping protein 1 (CTC1), the human orthologue of yeast STN1 telomerase capping complex subunit (STN1), and human ortholog of yeast TEN1 telomerase capping complex subunit (TEN1) (CST) complex has been shown to inhibit telomerase activity at late S/G₂ by competing with TPP1–POT1–telomerase complex in binding to telomeric DNA (39). S111 phosphorylation may therefore function as a molecular switch in balancing telomere-associated CST and telomerase activity for telomere length control.

Our study also provides a structural basis for telomerase recruitment and regulation. Although TPP1 shows basal interaction with telomerase throughout the cell cycle, phosphorylation of S111 specifically increases telomerase association and activity with TPP1 at the S to G₂/M phases. Whether this increase of telomerase activity was due to higher processivity remains to be determined. We noted that several conserved residues including D166, E168, E169, E171, and E215 within the telomerase-interacting domain of TPP1 are negatively charged. S111 is located in proximity to this telomerase-interacting domain, and its phosphorylation is expected to generate a more negatively charged surface, which may enhance telomerase association or activity (Fig. 4D). Alternatively, S111 phosphorylation may recruit additional molecules to stimulate telomerase. Based on our findings, small molecules may be designed that can specifically interfere with TPP1–telomerase interaction, thereby inhibiting the growth of telomerase-positive cancer cells.

Materials and Methods

Immunoprecipitation, Western Blotting, and Antibodies.

Coimmunoprecipitation studies were performed as described (40). Cells were lysed in buffer containing 20 mM Tris⋅HCl at pH 8.0, 1 mM EDTA, 100 mM NaCl, 0.5% Nonidet P-40, 1 mM DTT, and a proteinase inhibitor mixture. The lysates were then used for immunoprecipitations with appropriate antibodies for SDS/PAGE and Western blotting.

Rabbit polyclonal pS111Ab was generated by using a phosphopeptide based on human TPP1 sequences surrounding S111. To test the pS111Ab antibody, we generated HeLa cells expressing empty vector, or Flag-tagged wild-type or phosphorylation mutants (S111A, 7A, and 8A) of TPP1, and used these cells for immunoprecipitation experiments in the presence of phosphatase inhibitor mixture III (Sigma) with anti-Flag M2 resin. The samples were then treated with calf intestine phosphatase (41) (20 units in 30 μL of reaction) for 60 min at 37 °C in phosphatase buffer (100 mM NaCl, 50 mM Tris⋅HCl, 10 mM MgCl₂, and 1 mM DTT at pH 7.9) before Western blot analysis with the pS111Ab or anti-Flag antibody.

See more antibodies used for Western blotting and immunostaining in SI Materials and Methods.

Large-Scale Immunoprecipitation and Mass Spectrometry Sequencing.

For phosphopeptide identification, large-scale immunoprecipitation (IP) of Flag-tagged TPP1 followed by mass spectrometry sequencing was essentially performed as described (40). Cycling HTC75 cells expressing Flag-TPP1 and those treated with nocodazole (100 ng/mL, 16 h) were harvested and used for IP with anti-Flag M2 antibody-conjugated agarose beads (Sigma).

Cell Synchronization and Cell Cycle Analysis.

Double thymidine block was used to synchronize cells at G₁/S boundary. HTC75 cells were treated with 2.5 mM thymidine for 12 h, trypsinized, and replated to allow release from treatment for 14 h, before being treated again with 2.5 mM thymidine for another 12 h. For nocodazole treatment, 100 ng/mL nocodazole was added to cells for 16 h before analysis. To synchronize HeLa cells at G₂/M phase, cells were first treated with 2.5 mM thymidine for 24 h, released from treatment for 3 h, and then treated with 100 ng/mL nocodazole for 12 h. At the end of nocodazole treatment, the plates were tapped and rinsed to collect the detached round mitotic cells. These cells were washed twice with PBS and replated to allow cells entry into cell cycle.

To determine cellular DNA content, 1 × 10⁶ cells were collected from each time point, washed with PBS, and fixed with 95% ethanol at room temperature for 30 min. The fixed cells were then washed with PBS, resuspended with 0.5 mL of propidium iodide staining solution (50 µg/mL propidium iodide in 1×PBS containing 0.2 mg/mL DNase-free RNase A at pH 7.4), and incubated at 37 °C for 30 min before flow cytometry analysis by using a LSRII analyzer (BD Biosciences).

Analysis of TPP1-Associated Telomerase Activity.

Cells (5–10× 10⁶) were lysed in 5× pellet volume of high-salt buffer (20 mM Hepes at pH 7.9, 0.42 mM KCl, 25% (vol/vol) glycerol, 0.2% Nonidet P-40, 0.1 mM EDTA, 1 mM DTT, and protease inhibitors) on ice for 30 min. The lysates were then diluted with 5× volume of low-salt buffer (20 mM Hepes at pH 7.9, 100 mM KCl, 25% (vol/vol) glycerol, 0.1 mM EDTA, 1 mM DTT, and protease inhibitors) and centrifuged at >18,000 × g for 10 min at 4 °C. The supernatant was incubated with anti-FLAG M2-agarose beads (Sigma) and the immunoprecipitated proteins were eluted in 150–180 μL of elution buffer (25 mM Tris⋅HCl at pH 7.4, 136 mM NaCl, 2.6 mM KCl, 1 mM MgCl₂, 1 mM EGTA, 10% (vol/vol) glycerol, 1 mM DTT, and protease and RNase inhibitors) and Flag peptides (200 µg/mL). The eluate was then diluted two- to fivefold before being used for real-time quantitative PCR-based TRAP assay (42). Each 25 μL of real-time quantitative PCR-based TRAP reaction contained 2 μL of the eluted proteins, 100 ng each of TS primer (5′-AATCCGTCGAGCAGAGTT-3′) and ACX primer (5′-GCGCGGCTTACCCTTACCCTTACCCTAACC-3′), and 1 mM EGTA in SYBR Green PCR Master Mix (Applied Biosystems). The reaction mixtures were incubated at 30 °C for 30 min and then PCR amplified (40 cycles of 95 °C for 15 s and 60 °C for 60 s) by using an ABI StepOnePlus Real-Time PCR System (Applied Biosystems).

TRF Assay.

Analysis of HTC75 cells stably expressing control and FLAG-tagged wild-type and mutant TPP1 proteins by using the TRF assay was done as described (13, 43).

Structural Modeling of Human TPP1 OB Fold.

The phosphate group of S111 was attached by using Coot (44) with its modeling feature, assuming no major structural perturbation caused by the phosphorylation.