Full text

Turn on search term navigation

ARTICLE

Received 27 Oct 2014 | Accepted 27 May 2015 | Published 9 Jul 2015

DOI: 10.1038/ncomms8643 OPEN

Telomeric G-quadruplexes are a substrate and site of localization for human telomerase

Aaron L. Moye1,*, Karina C. Porter1,*, Scott B. Cohen1,*, Tram Phan2, Katherine G. Zyner1, Natsuki Sasaki1, George O. Lovrecz2, Jennifer L. Beck3 & Tracy M. Bryan1

It has been hypothesized that G-quadruplexes can sequester the 30 end of the telomere and prevent it from being extended by telomerase. Here we purify and characterize stable, conformationally homogenous human telomeric G-quadruplexes, and demonstrate that human telomerase is able to extend parallel, intermolecular conformations in vitro. These G-quadruplexes align correctly with the RNA template of telomerase, demonstrating that at least partial G-quadruplex resolution is required. A highly puried preparation of human telomerase retains this extension ability, establishing that the core telomerase enzyme complex is sufcient for partial G-quadruplex resolution and extension. The parallel-specic G-quadruplex ligand N-methyl mesoporphyrin IX (NMM) causes an increase in telomeric G-quadruplexes, and we show that telomerase colocalizes with a subset of telomeric G-quadruplexes in vivo. The ability of telomerase to partially unwind, extend and localize to these structures implies that parallel telomeric G-quadruplexes may play an important biological role.

1 Childrens Medical Research Institute, University of Sydney, 214 Hawkesbury Road, Westmead, New South Wales 2145, Australia. 2 Commonwealth Scientic and Industrial Research Organisation, Manufacturing Flagship, 343 Royal Parade, Parkville, Victoria 3052, Australia. 3 School of Chemistry, University of Wollongong, Northelds Avenue, Wollongong, New South Wales 2522, Australia. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed to T.M.B. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

Human chromosomes terminate with B510 kb of the telomeric DNA repeat TTAGGG1,2. The shortening of telomeres to a critical length is a signal for cellular

senescence3,4. Telomere shortening is counteracted by telomerase, the telomere-specic reverse transcriptase that contains an RNA template to direct the addition of telomeric DNA by the catalytic telomerase reverse transcriptase (TERT) component57. Telomerase is upregulated in Z85% of tumours, providing a telomere maintenance mechanism that contributes to cellular immortalization and tumour progression8,9.

Telomeric DNA can fold into compact G-quadruplex structures involving the interaction of four guanine bases in a square planar arrangement stabilized by central cations10. G-quadruplexes can form from one, two or four strands of DNA in a parallel or antiparallel orientation11. Over 375,000 sequences with the potential to form G-quadruplexes have been identied within the human genome12. There is increasing evidence supporting in vivo functions for DNA and RNA G-quadruplexes, including telomere protection and involvement in transcription, translation and splicing1316. Direct evidence for the existence of G-quadruplexes at telomeres was obtained using specic antibodies against G-quadruplexes in the telomeric DNA of the ciliate Stylonychia lemnae17. Recently, a G-quadruplexspecic antibody has also been used to directly visualize G-quadruplexes throughout the human genome, a portion of which is located at telomeres18. There are a large number of small-molecule ligands that have specicity for G-quadruplexes, with many of them causing telomere dysfunction in vivo and inhibiting telomerase activity in vitro1921.

The rst report describing the inability of telomerase to extend G-quadruplexes, using the ciliate Oxytricha nova, did not differentiate between specic conformations as substrates22. In 2006, we demonstrated using puried G-quadruplex structures that some conformations can be extended by telomerase from the ciliated protozoa Tetrahymena thermophila and Euplotes aediculatus: parallel-stranded intermolecular G-quadruplexes were substrates for ciliate telomerase, whereas antiparallel intramolecular G-quadruplexes were not23. Telomere biology is very different between ciliated protozoa and humans, and the telomerase ribonucleoproteins from the two groups differ in key properties such as the ability to extend non-telomeric DNA, the mechanism of template boundary denition and the components and multimeric state of the enzyme complex2428. The possibility therefore existed that the ability to extend parallel G-quadruplexes is a ciliate-specic property of telomerase. Given the intense interest in telomerase inhibition by G-quadruplex stabilizers as a potential cancer therapeutic20,21, and the polymorphism of human telomeric G-quadruplexes11, we considered it important to determine the ability of human telomerase to extend dened conformations of G-quadruplex.

In this study, we establish that the ability to extend parallel intermolecular G-quadruplexes is conserved among evolutionarily distant species. We characterize puried, stable, conformationally homogenous human telomeric G-quadruplexes

and demonstrate that human telomerase is able to recognize and partially unwind parallel intermolecular conformations and extend the unwound 30 end in vitro. Treatment of immortal human cells with the parallel-specic G-quadruplex ligand N-methyl mesoporphyrin IX results in a signicant increase in G-quadruplexes present at telomeres in vivo, indicating the potential for formation of a parallel G-quadruplex structure at human telomeres. We show that human telomerase localizes to a subset of telomeres that contain G-quadruplexes in vivo.

ResultsFolding and characterization of parallel G-quadruplexes.To test the ability of human telomerase to extend G-quadruplex DNA, we isolated and characterized parallel, intermolecular G-quadruplexes, since this is the conformation extended by ciliate telomerase23. To be certain that telomerase extends the G-quadruplex rather than its unfolded linear counterpart, it is necessary to work with structures with very slow unfolding rates under the conditions of a telomerase activity assay (30 C, 1 h).Human telomeres have one less contiguous guanine residue than their ciliate counterparts (TTAGGG versus TTGGGG), which renders the corresponding G-quadruplexes less stable29.It is possible, however, to form stable human telomeric G-quadruplexes using cations such as Sr2 , rather than K or

Na (refs 30,31). Tetramolecular G-quadruplexes formed from short oligonucleotides also tend to be relatively stable29,32.

The oligonucleotide 7GGT (Table 1) formed a tetrameric G-quadruplex, [7GGT]4, when annealed in buffer containing 150 mM KCl, and gave a single band upon native gel electrophoresis. This band was puried from the native gel and remained 498% pure after purication (Fig. 1a). The circular dichroism (CD) spectrum of [7GGT]4 gave a positive signal at 260 nm and a negative signal at 240 nm, characteristic of a parallel G-quadruplex (Fig. 1b)33. The ultraviolet thermal difference spectrum (TDS)34 of [7GGT]4 showed a negative peak at 295 nm (Supplementary Fig. 1a), supporting the assignment of a G-quadruplex structure29,32.

The oligonucleotide 22GGG (Table 1) gave rise to two bands on a native gel when annealed in 2.5 mM Sr2 (Fig. 1c). The migration of the upper and lower bands was consistent with formation of intermolecular and intramolecular structures, respectively10. Each G-quadruplex band remained Z95% pure after gel purication (Fig. 1c); no linear 22GGG was observed after purication. CD analysis of the upper band was indicative of a parallel G-quadruplex structure (Fig. 1d), whereas the lower band gave a maximum absorbance at 295 nm and a minimum at 260 nm, consistent with an antiparallel structure (Fig. 1d). The TDS spectra for both the upper and lower bands of 22GGG conrmed G-quadruplex structures (Supplementary Fig. 1b,c).

Electrospray ionization mass spectrometry (ESI-MS) was used to determine G-quadruplex stoichiometry35,36. The puried 22GGG lower band gave predominant ions that were assigned to unimolecular species, consistent with the inclusion of an

Table 1 | Oligonucleotides used and G-quadruplex structures characterized in this study.

Nomenclature Sequence (5030) Length (nt)

G4 nomenclature

Annealing cation

Strand orientation

Strand stoichiometry

T (C)

7GGT TTAGGGT 7 [7GGT]4 150 mM K Parallel Tetramer 68 22GGG AGGG(TTAGGG)3 22 [22GGG]2 2.5 mM Sr2 Parallel Dimer 490 22GGG AGGG(TTAGGG)3 22 [22GGG]1 2.5 mM Sr2 Antiparallel Intramolecular 68 9C CCCTAACCC 9

T15 TTTTTTTTTTTTTTT 15 T25 TTTTTTTTTTTTTTTTTTTTTTTTT 25

2 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643 ARTICLE

[7GGT] 4

7GGT

T 25

T 15

Molar ellipticity

(deg.cm2 dmol1 )

3.E+052.E+051.E+05

1.E+05 2.E+04

0.E+00

240

260

280

300

320

Wavelength (nm)

[22GGG] 2

Unpurified

[22GGG] 1

T 15

T 25

22GGG upper

22GGG lower

6.E+054.E+052.E+040.E+04 2.E+04

Molar ellipticity

(deg.cm2 dmol1 )

240

260

280

300

320

Wavelength (nm)

e f

100

1,200 1,400 1,600 1,800 2,000 1,200 1,400 1,600 1,800 2,000

m/z

Figure 1 | Characterization of gel-puried telomeric G-quadruplexes.(a) Native gel electrophoresis of puried 32P-labelled [7GGT]4 in 150 mM

K . Lane 1: [7GGT]4 after gel purication. Lane 2: unfolded 7GGT. Lanes 3 and 4: T25 and T15 unstructured molecular weight (MW) markers.(b) CD spectrum of gel-puried [7GGT]4 in 150 mM K . (c) Native gel electrophoresis of G-quadruplexes formed from 22GGG in 2.5 mM Sr2 , stained in SYBR Gold. Lane 1: unpuried folded 22GGG. Lane 2: post-purication folded 22GGG upper band, [22GGG]2. Lane 3: post-purication 22GGG lower band, [22GGG]1. Lanes 4 and 5: unstructured MW markers T25 and T15. (d) CD spectra of gel-puried 22GGG G-quadruplexes in2.5 mM Sr2 . (e) Negative ion ESI mass spectra of the lower band of folded 22GGG and (f) the upper band of folded 22GGG. K [M-5H]5 ;

B [M-4H]4 ; J [M Sr2 -7H]5 ; ~ [M Sr2 -6H]4 ;

[2M 3Sr2 -13H]7 ; & [2M 3Sr2 -14H]8 .

Sr2 ion within an intramolecular G-quadruplex (Fig. 1e; Supplementary Table 1). A lower-abundance ion can be assigned as the dimeric G-quadruplex species [2M 3Sr2-

13H]7 (where M is neutral, unimolecular 22GGG), consistent with the observation of some reformation of the upper parallel

G-quadruplex after purication. Together, the migration pattern on native gel electrophoresis, ESI-MS and CD analysis indicate that the lower band of 22GGG annealed in Sr2 is an intramolecular antiparallel G-quadruplex, which will henceforth be referred to as [22GGG]1.

The predominant ions observed from the puried upper band of 22GGG represented bimolecular DNA along with a unimolecular G-quadruplex (Fig. 1f; Supplementary Table 1); this is consistent with the observation of some upper band converting to the intramolecular G-quadruplex after purication. The ESI mass spectrum, combined with CD and TDS analysis and native gel electrophoresis, is consistent with a parallel, dimeric G-quad-ruplex, here called [22GGG]2.

Absence of spontaneous unfolding of G-quadruplexes. The relative stabilities of the puried G-quadruplexes were initially assessed by determining their melting temperatures. Since inter-molecular G-quadruplexes reform very slowly and are therefore

not at equilibrium during the measurement, we refer to this value as T rather than Tm, which is the true thermodynamic parameter at equilibrium29. The ultraviolet melting curve of [7GGT]4 gave a T of 68 C (measured at a heating rate of 1 C min 1; Supplementary Fig. 1d; Table 1).

To determine the amount of [7GGT]4 that unfolds during the course of a telomerase activity assay (30 C, 1 h), we employed a complementary C-strand trap assay. A 10-fold excess of C-rich complementary strand (Table 1) was added to the G-quadruplex in the same buffer used for telomerase activity assays (described below), and the extent of WatsonCrick duplex formation was monitored over 4 h by non-denaturing gel electrophoresis37,38 (Fig. 2a). Control experiments established that the result was unaffected by increasing C-strand concentrations, up to 100-fold over G-quadruplex (Supplementary Fig. 2a), demonstrating that the rate of WatsonCrick duplex formation reects the rate of G-quadruplex unfolding. After 1 h at 30 C, B1% of the

G-quadruplex had hybridized to the complementary strand, with B34% hybridized after 4 h (Fig. 2a,b). As a further test of the stability of [7GGT]4, the G-quadruplex was treated with phosphodiesterase 1 (P1), which cleaves single-stranded DNA exonucleolytically at the 30 end. Under conditions identical to the telomerase activity assays below (30 C for 1 h), all of the P1-treated linear 7GGT was digested while no [7GGT]4 digestion was detected (Fig. 2c). The 7-nt oligonucleotide remained intact after digestion, demonstrating that even the single thymidine at the 30 end of each of the four strands is participating in tertiary structure and being protected from digestion. Overall, these data demonstrate that the G-quadruplex [7GGT]4 is stable under the conditions of the telomerase activity assay.

The T values of [22GGG]2 and [2GGG]1 were 490 and 68 C, respectively, when measured at 1 C min 1 (Supplementary Fig. 1e,f; Table 1). The high melting temperature of the dimeric species indicates that it is an exceptionally stable structure. An identical complementary C-strand experiment was performed with [22GGG]2 in 2.5 mM Sr2 buffer. After 4 h at 30 C, there was no observable duplex formation, indicating that the [22GGG]2 G-quadruplex does not unfold under our assay conditions (Fig. 2d). This was conrmed by digestion with P1 at 30 C for 1 h; no digestion of [22GGG]2 was observed (Fig. 2e,f, right lanes). It should be noted that any [22GGG]2 that did unfold over days or weeks in storage rapidly formed [22GGG]1, with no linear 22GGG oligonucleotide observed (Fig. 1c).

Intramolecular G-quadruplexes often display rapid folding and unfolding rates23,39. Under the conditions used in this study, it was not possible to obtain a native gel or CD spectrum that corresponded to linear, unfolded 22GGG in the presence of2.5 mM SrCl2. When incubated with P1, only B30% of linear 22GGG in 2.5 mM SrCl2 was digested (Fig. 2e,f, middle lanes). This suggests that linear 22GGG in the presence of Sr2 is in rapid equilibrium with [22GGG]1. To conrm this, we locked linear 22GGG into an unfolded conformation by pre-binding it with the 9C complementary oligonucleotide (Supplementary Fig. 2b), as previously described39. After digestion with P1, only 10% of the DNA remained undigested (Fig. 2e,f, left lanes). These data demonstrate that the P1 exonuclease was functional in this experiment, and that linear 22GGG and intramolecular [22GGG]1 G-quadruplex are in rapid equilibrium when in the presence of 2.5 mM Sr2 .

Parallel G-quadruplexes are substrates of human telomerase. Telomerase activity assays were performed with the G-quadruplexes characterized above to determine whether parallel, intermolecular G-quadruplexes are a substrate for human telomerase. Telomerase was prepared by overexpressing its

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

a c

Time (min)

T15

7GGT

WC duplex

[7GGT] 4

1 5 15

30 60 120

240

[7GGT] 4

WC duplex

7GGT [7GGT]4

[7GGT] 4remaining (%)

100

Control

P1 +P1 P1 +P1

60 120

180 240

Time (min)

1 2 3 4 5 6 7 8 9 10 11 12 13

d f

Time (min)

22GGG

WC duplex

[22GGG] 2

Control

22GGG +9C

P1 +P1 P1 +P1 P1 +P1

[22GGG]2

T25

1 2 3 4 5 6 7 8

120

240

Undigested (%)

100

P1: + +

0 +

10 11 12

22GGG +9C

22GGG [22GGG]2

Figure 2 | Stability of telomeric G-quadruplexes. (a) [7GGT]4 complementary strand trap assay: 50-end-labelled [7GGT]4 in 150 mM K , incubated with 10-fold excess complementary strand 9C at 30 C and electrophoresed on a 12% non-denaturing polyacrylamide gel. Lane 1: unstructured molecular weight (MW) marker T15. Lane 2: denatured 7GGT. Lane 3: pre-annealed WatsonCrick (WC) duplex. Lane 4: gel-puried [7GGT]4 G-quadruplex. Lanes 511: [7GGT]4 incubated with 9C at 30 C for the indicated times. Lane 12: [7GGT]4. Lane 13: pre-annealed WC duplex. Samples were loaded at the times indicated, with the gel running between loadings, so later samples underwent shorter electrophoresis times. The last three lanes were loaded simultaneously. (b) Proportion of [7GGT]4 remaining over time at 30 C. Error bars represent s.d. of the mean of three independent experiments.

(c) P1 digestion of 7GGT and [7GGT]4 electrophoresed on a 12% denaturing polyacrylamide gel; digestions were performed in triplicate. Control lane: end-labelled 7GGT. (d) [22GGG]2 complementary strand trap assay. Unlabelled [22GGG]2 in 2.5 mM Sr2 , incubated with 10-fold excess complementary strand 9C at 30 C, electrophoresed on a 16% non-denaturing polyacrylamide gel and stained in SYBR Gold. Lane 1: unstructured MW marker T25. Lane 2:

pre-annealed WC duplex. Lane 3: gel-puried [22GGG]2. Lanes 411: G-quadruplex [22GGG]2 incubated with 9C at 30 C for the indicated times. Lane 12: gel-puried [22GGG]2. (e) P1 digestion of 22GGG hybridized to complementary strand 9C, linear 22GGG and [22GGG]2, electrophoresed on a 12%

denaturing polyacrylamide gel; digestions were performed in triplicate. Control lane: end-labelled 22GGG. (f) Data from e, quantied. Error bars represent s.d. of the mean of triplicates. LC indicates a 32P-labelled 100-nt oligonucleotide used as a control for loading and recovery.

components in 293T cells40 and immunopurifying the assembled complex with an hTERT antibody41. The G-quadruplex [7GGT]4 was readily extended by human telomerase (Fig. 3a): the MichaelisMenton constant (Km) for 7GGT was 10020 nM and that for [7GGT]4 was 500120 nM (means.d.; n 46).

Notably, both substrates supported equal amounts of catalytic activity, with a relative Vmax (folded/linear) of 0.970.07 (means.d.; P 0.3, one-sample t-test; n 7).

There is a possibility that the one-step telomerase immunopurication method used in these experiments could allow co-immunopurication of other proteins that may selectively unwind the parallel G-quadruplexes, allowing the observed extension. To exclude this possibility, we puried human telomerase further with a four-step procedure involving immunopurication, capture onto an immobilized telomeric DNA substrate, release of active enzyme and sedimentation over a glycerol gradient42 (Fig. 3b); the specic activity of telomerase on an 18-nt telomeric substrate does not change over the course of purication (Supplementary Fig. 3). The purity of this telomerase was conrmed on a silver-stained SDSpolyacrylamide gel electrophoresis (PAGE) gel (Fig. 3b); the only visible protein bands correspond to the sizes of the core telomerase components hTERT and dyskerin42. The G-quadruplex helicase DHX36 (also known as G4R1 or RHAU) has been shown to associate with a portion of active telomerase in immortal human cells43,44;

we conrmed this association in our 293T system by immunoblot analysis for DHX36 on one-step immunopuried telomerase (Fig. 3c). However, our four-step purication procedure resulted in loss of DHX36, conrmed by immunoblot of an equal molar amount of highly puried telomerase (GG in Fig. 3c).

Telomerase activity assays were performed with this highly puried telomerase alongside one-step immunopuried telomerase. The relative activity using [7GGT]4 G-quadruplex and 7GGT linear DNA as substrates was the same for highly puried as for one-step puried telomerase (Fig. 3d). This conrms that the core telomerase enzyme complex is sufcient for extension of [7GGT]4.

The dimeric parallel G-quadruplex [22GGG]2 was also readily extended by immunopuried human telomerase, with a Km of

5310 nM (Fig. 4a; means.d.; n 4), under conditions in

which we had demonstrated that this G-quadruplex is extremely stable. [22GGG]2 was also extended by the highly puried telomerase preparation (Fig. 4b). Thus, the core human telomerase complex is capable of extending multiple parallel, intermolecular G-quadruplexes.

Obtaining reproducibly quantitative telomerase activity assays for linear 22GGG and [22GGG]1 was confounded by the rapid interconversion between the folded and unfolded forms. It is well established that intramolecular, antiparallel G-quadruplexes are not good substrates for either ciliate or human telomerase22,23,39.

4 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643 ARTICLE

No DNA

7GGT [7GGT]4

7GGT

[7GGT] 4

kDa

160

120

100

0.01

0.05

0.1

0.25

0.5

0.75

1.5

0.01

0.05

0.1

0.25

0.5

0.75

1.5

220

M GG

GG GG GG

IP IP IP

hTERT

Dyskerin

kDa

120

100

220

M IP GG

IB: DHX36

hTR

Figure 3 | Extension of 7GGT and [7GGT]4 by human telomerase. (a) Telomerase activity assays with immunopuried telomerase and the indicated concentrations of either 7GGT or [7GGT]4. The concentrations of [7GGT]4 refer to concentrations of the assembled, four-stranded G-quadruplex.

(b) Silver-stained SDSPAGE gel of highly puried telomerase (GG). (c) Immunoblot for DHX36 with immunopuried (IP) or highly puried (GG) telomerase. The bottom panel shows a dot-blot northern for hTR, demonstrating equal loading of telomerase in the gel. (d) Telomerase activity assays with 2 mM 7GGTor [7GGT]4 with immunopuried (IP) or highly puried (GG) telomerase, using equal amounts of each enzyme; reactions were performed in duplicate. For a and d, the asterisk indicates the rst visible addition product (n 3); LC indicates a 32P-labelled 100-nt oligonucleotide used as a control

for loading and recovery.

To conrm that this is also the case for the [22GGG]1 intramolecular G-quadruplex folded in Sr2 , we again locked

linear 22GGG in Sr2 into the unfolded form by hybridizing it with the complementary DNA oligonucleotide 9C39, which resulted in a twofold recovery of activity (Fig. 4c,d). In contrast, addition of the complementary strand to the [22GGG]2

G-quadruplex did not affect activity (Fig. 4c,d). These data support our hypothesis that unfolded 22GGG is in equilibrium between linear and antiparallel G-quadruplex conformations, and conrm that the antiparallel [22GGG]1 G-quadruplex is a less-favourable substrate for human telomerase than its linear counterpart. Taken together, these data demonstrate that multiple parallel, intermolecular G-quadruplexes are substrates for human telomerase, while an antiparallel, intramolecular G-quadruplex is not, despite the much higher stability of the former G-quadruplexes. We have also demonstrated that the core human telomerase enzyme complex is sufcient for parallel G-quadruplex extension.

Telomerase partially unwinds parallel G-quadruplexes. We have demonstrated that, for Tetrahymena telomerase, parallel G-quadruplex substrates increase the Km of telomerase for the incoming dTTP nucleotide, presumably by perturbation of the nucleotide-binding site by the wider G-quadruplex substrate37. The Km of human telomerase for dTTP was determined in the presence of saturating [7GGT]4 G-quadruplex and 7GGT linear DNA at a range of dTTP concentrations (Fig. 5a). The Km of

telomerase for dTTP increased approximately vefold in the presence of the G-quadruplex substrate, from 2.80.5 mM for linear 7GGT to 155 mM with [7GGT]4 (means.d.; P 0.0136,

unpaired two-tailed t-test; n 3). This change in Km provides

evidence that the observed extension is occurring with two conformationally distinct substrates.

We demonstrated the maintenance of G-quadruplex structure following telomerase extension directly, by visualizing telomerase

extension products by native gel electrophoresis. A mixture of unlabelled [22GGG]1 and [22GGG]2 was extended by telomerase in the presence of [a-32P]-dTTP (Fig. 5b,c). Under native gel electrophoresis conditions, a single band was present following telomerase extension (Fig. 5c, lane 3) that aligns with an authentic

32P-labelled [22GGG]2 marker (Fig. 5c, lane 2). These data provide evidence that the [22GGG]2 parallel G-quadruplex structure is maintained following extension by telomerase. Direct observation of [22GGG]2 specically labelled through extension by telomerase provides unequivocal evidence that G-quadruplex DNA must be the substrate utilized by telomerase.

The telomerase RNA template hybridizes to telomeric DNA to enable telomere extension with the correct alignment. Ciliate telomerase can perform extension of non-telomeric substrates without any complementarity between the DNA primer and RNA template, using a specic template position as the default for the initiation of extension26. We wished to test whether the extension of parallel G-quadruplexes is preceded by canonical template hybridization (Fig. 5b). We performed telomerase activity assays in the presence of different combinations of nucleotides, for both [7GGT]4 and [22GGG]2 (Fig. 5d,e). A standard processive reaction including [a-32P]-dGTP, dATP and dTTP provided a marker for band positions, as the major pause site in a processive reaction is at the end of the template (products ending in TAG; see labelling of gel in Fig. 5d,e). In the presence of only [a-32P]-dTTP and dATP, we anticipated the addition of two nucleotides (TA) and three nucleotides (TTA) to primers 7GGT and 22GGG, respectively, if the terminal 30 nucleotides are correctly aligning with the RNA template (Fig. 5b). Because addition of ddGTP results in termination of DNA strand elongation, we anticipated the addition of three nucleotides (TAG) and four nucleotides (TTAG) to 7GGT and 22GGG, respectively, in the presence of [a-32P]-dTTP, dATP and ddGTP.

In all cases, the patterns of extension of parallel G-quadruplexes [7GGT]4 and [22GGG]2 are identical to their respective linear

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

a b

[22GGG]2

500

22GGG

No DNA

22GGG

[22GGG] 2

250

100

500

250

100

nM [Mol]

[22GGG]1 [22GGG]2

C-strand

+ + + +

Relative activity (%)

100

C-strand +

[22GGG]1 [22GGG]2

Figure 4 | Extension of linear 22GGG and [22GGG]2, but not [22GGG]1, by human telomerase. (a) Telomerase activity assays with immunopuried telomerase and the indicated concentrations of either linear 22GGG or [22GGG]2. (b) Telomerase activity assays with 1 mM linear 22GGG or [22GGG]2 with immunopuried (IP) or highly puried (GG) telomerase, using equal amounts of each enzyme. (c) Telomerase activity assays with immunopuried telomerase and 1 mM [22GGG]1 and [22GGG]2, with or without complementary strand 9C as indicated; reactions were performed in duplicate.

(d) Quantication of c normalized against 22GGG pre-annealed with 9C. Error bars represent s.d. of the mean of three independent experiments.

In all panels, the asterisk indicates the rst visible addition product (n 4), and LC indicates a 32P-labelled 100-nt oligonucleotide used as a control

for loading and recovery.

controls (Fig. 5d,e) and correspond with expected nucleotide addition patterns (Fig. 5b). For canonical hybridization to occur, at least a portion of the parallel G-quadruplex structure must be unwound. As native gel electrophoresis demonstrated alignment of the extended parallel G-quadruplex with an identical

32P-labelled marker (Fig. 5c), we conclude that parallel G-quadruplexes are partially unwound to allow hybridization of the telomerase template with the 30 end of the DNA.

Telomerase localizes to G-quadruplex-containing telomeres. To investigate the presence of parallel telomeric G-quadruplexes in cells, we performed immunouorescence on human embryonic kidney 293T cells using a G-quadruplex-specic antibody18. Fixed cells were subjected to cytoplasmic extraction prior to immunouorescence to remove RNA G-quadruplexes; a DNase-treated control demonstrates the specicity of the antibody for DNA G-quadruplexes under these conditions (Fig. 6a, top right). A subset of telomeres colocalized with G-quadruplex foci (Fig. 6a, top panel), as previously demonstrated18. Treatment of the cells with the parallel-specic G-quadruplex porphyrin ligand N-methyl mesoporphyrin IX4547 (Supplementary Fig. 4a) during

mid-S phase of the cell cycle resulted in an increase in the number of G-quadruplextelomere colocalizations per cell (Fig. 6a,b), without affecting progression of the cells into mid-S phase (Supplementary Fig. 4b). These data provide evidence that parallel G-quadruplexes can form at human telomeres in vivo.

Since human telomerase can extend parallel G-quadruplexes in vitro, we asked whether telomeric G-quadruplexes are a site of localization for human telomerase in vivo. 293T cells were synchronized in mid-S phase, which is the peak of telomerase recruitment to telomeres48. Using G-quadruplex immunouorescence, combined with uorescence in situ hybridization (FISH) for telomerase RNA (hTR) and telomeres49, we observed the simultaneous colocalization of endogenous telomerase, G-quadruplexes and telomeres (Fig. 6c; Supplementary Fig. 5). Control experiments demonstrate that depletion of dyskerin, a component of the core telomerase complex42, markedly reduces hTR foci, verifying the specicity of the FISH signal (Supplementary Fig. 6). G-quadruplex foci colocalized with B17% of total hTR-telomere foci (Fig. 6d). These results reveal that human telomerase can localize to telomeres at which G-quadruplexes are present.

6 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643 ARTICLE

hTR template

7GGT [7GGT]4

mM dTTP

1.25

2.5

12.5

7.25

3.75

100

1.25

12.5

7.25

3.75

2.5

100

Labelled markers

Extension reaction

[22GGG] 1

[22GGG] 1+

[22GGG] 2

[22GGG] 1+

[22GGG] 2

Marker

+TT

1 3

d e

22GGG [22GGG]2

1 2 3 4 5 6

+TTA

7GGT

[7GGT]4

T,A,ddG

T,A,G

T,A,ddG

22GGG

T 25

T,A,ddG

T,A,G

T,A

T,A,ddG

T,A,G

T,A

T,A,G

+TAG(GGTTAG)2

+TAGGGTTAG

+TTAG(GGTTAG)2

+TTAGGGTTAG

+TAG

+TTAG

+TA

1 2 3 4 5 6 7 8

Figure 5 | Telomerase partially unwinds and extends [7GGT]4 and [22GGG]2. (a) Telomerase activity assays with immunopuried telomerase, using 1 mM 7GGT or [7GGT]4 and the indicated concentrations of dTTP. The asterisk indicates the rst visible addition product (n 3). (b) Expected telomerase

products upon alignment of 7GGTor 22GGG with the RNA template in the presence of different combinations of nucleotides. (c) Telomerase activity assay followed by native gel electrophoresis. Lane 1: 50-end-labelled [22GGG]1. Lane 2: mixture of 50-end-labelled [22GGG]1 and [22GGG]2, demonstrating separation of the two species. Lane 3: products of unlabelled [22GGG]1 and [22GGG]2 following telomerase extension in the presence of [a-32P]-dTTP.

Unstructured marker added for orientation purposes. (d) Templated addition of nucleotides to 7GGT and [7GGT]4. G-quadruplexes and their linear counterparts (2 mM) were incubated with telomerase in the presence of: lanes 1 and 5: 32P-dTTP and dATP; lanes 2 and 6: 32P-dTTP, dATP and ddGTP;

lanes 3 and 7: 32P-dTTP, dATP and dGTP; lanes 4 and 8: dTTP, dATP and 32P-dGTP, and products electrophoresed on a denaturing 12% acrylamide gel. Migration position of different products is shown. (e) Templated addition of nucleotides to 22GGG and [22GGG]2. G-quadruplexes and their linear counterparts (2 mM) were incubated with telomerase in the presence of: lanes 1 and 4: 32P-dTTP and ddATP; lanes 2 and 5: 32P-dTTP, dATP and ddGTP;

lanes 3 and 6: dTTP, dATP and 32P-dGTP, and products electrophoresed on a denaturing 12% acrylamide gel. End-labelled T25 and 22GGG markers are on the left of the gel. Migration position of different products is shown. In all panels, LC indicates a 32P-labelled 100-nt oligonucleotide used as a control for loading and recovery.

DiscussionIt has been hypothesized that G-quadruplexes can sequester the 30 end of the telomere and prevent it from being extended by telomerase22,39. Here, we demonstrate that parallel intermolecular telomeric G-quadruplexes are partially unwound and robustly extended by human telomerase. We used highly puried telomerase to show that the core telomerase enzyme complex

has the ability to partially unwind the G-quadruplexes prior to their extension.

The apparent afnity of telomerase for the incoming dTTP nucleotide was signicantly reduced when extending [7GGT]4 compared with linear 7GGT, suggesting a change in the conformation of (or near) the active site of telomerase to accommodate the wider G-quadruplex structure. This difference

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

TTAGGG G4 Merge

hTR

DAPI

TTAGGG G4 DAPI

DNAse (control)

No treatment DMSO NMM

Untreated

****

TelomereG4

colocalizations per cell

DMSO

NMM

Average localizations

per cell

Total hTR+

telomere

hTR+

0 telomere+G4

1.3

1.4

TTAGGG G4 Merge

1.2

1.0

0.8

No treatment

0.6

0.4

0.22

0.2

Figure 6 | G-quadruplex-containing telomeres are a site of localization for human telomerase. (a) Immunouorescence with G-quadruplex antibody BG4 together with telomere FISH, in mid-S-phase HEK 293T cells with no treatment (top), vehicle (dimethyl sulfoxide (DMSO), middle) or parallel-specic ligand N-methyl mesoporphyrin IX (NMM) (bottom). DNAse treatment is included as a G-quadruplex antibody control (top right).(b) Quantication of number of telomere-G-quadruplex colocalizations per cell from a. One hundred nuclei were counted per condition in two (DMSO) or three (untreated and NMM) independent experiments. Error bars represent the s.e.m. Statistical signicance was calculated using an unpaired t-test; ****Po0.0001 (c) Immunouorescence with G-quadruplex antibody BG4 together with FISH for telomeres and telomerase RNA (hTR). (d) Quantication of hTRtelomereBG4 trilocalization frequency, relative to telomerasetelomere localization events. Hundred nuclei were counted in each of the three independent experiments. Error bars represent the s.e.m. Scale bars, 10 mm. NS, not signicant.

G-quadruplex extension by telomerase

of Km with different conformations of DNA substrate demonstrates that the parallel G-quadruplex is a substrate for telomerase while still at least partially structured, an observation we conrmed using native gel electrophoresis of a parallel G-quadruplex following extension (Fig. 5c). Nevertheless, the telomerase extension patterns demonstrate correct alignment of the 30 end of the DNA substrate with the RNA template, indicating that parallel G-quadruplexes are partially unwound by or invaded by telomerase. Neither of the parallel G-quadruplexes in this study demonstrated any hybridization to a complementary DNA oligonucleotide or digestion by a nuclease under the conditions of the telomerase activity assays, arguing against a model of spontaneous transient unfolding of the DNA 30 ends followed by telomerase extension. Given the exceptional stability of the [22GGG]2 G-quadruplex, in particular, these data support an extension model where the parallel G-quadruplex structure is bound and subsequently partially resolved by telomerase prior to extension of the DNA 30 end (Fig. 7). We are currently investigating the mechanism of this resolution, including the regions of telomerase responsible, given that neither hTERT nor dyskerin contains a known helicase domain. It is possible that telomerase uses a similar mechanism as that proposed for replication protein A, involving binding to the single-stranded loop regions between stacks of G-quartets50.

Currently, there are a large number of laboratories investigating the interactions of stabilizing ligands with G-quadruplexes. It is therefore crucial to determine the specicity of human

telomerase for the different conformations of G-quadruplexes, such that once the in vivo structure(s) of human telomeres is determined, suitable G-quadruplex-stabilizing ligands of appropriate specicity can be rationally selected. Although in vivo evidence for G-quadruplex structures is increasing, the structure-specic locations and functions of G-quadruplexes remain speculative. Our in vivo data indicate that parallel G-quadruplexes can form at human telomeres, and that telomeres containing a G-quadruplex are a site of localization for human telomerase. That telomerase is also able to partially unwind and

hTERT

hTR DKC hTR

DKC

hTERT

Telomerase binding, partial resolution and hTR hybridization

Figure 7 | Proposed model for telomerase action at telomeric G-quadruplexes. Telomerase binds to a parallel G-quadruplex structure. The core telomerase complex is then able to partially unwind the G-quadruplex allowing hybridization of the RNA template and DNA extension.

8 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643 ARTICLE

extend these structures, and the evolutionary conservation of this property in distantly related organisms such as Tetrahymena23, implies that telomeric DNA in a parallel G-quadruplex conformation may have a biological role. In support of this idea, it has been demonstrated that promotion of parallel telomeric G-quadruplexes by the Saccharomyces cerevisiae protein Est1p is essential for telomerase-mediated telomere elongation in that organism51. One potential biological role for telomeric G-quadruplexes in human cells may be during meiosis, when two pairs of sister chromatids are brought together; it has been suggested that parallel G-quadruplexes could be responsible for the correct alignment of the four chromatids52. The ability of telomerase to partially unwind and extend parallel G-quadruplexes potentially formed during meiosis may be necessary for the maintenance of germline telomere length. Alternatively, G-quadruplexes may be involved in the association of telomeres of sister chromatids during S phase, which is known to be necessary for telomerase to lengthen telomeres in human cells53. The data presented here provide evidence that subtypes of telomeric G-quadruplexes interact differentially with human telomerase, possibly reecting their different biological roles.

Methods

Oligonucleotide preparation. DNA oligonucleotides (Table 1) were purchased from Sigma Genosys in desalted form. All oligonucleotides were puried by electrophoresis on denaturing 20% polyacrylamide/8 M urea gels in 1 TBE buffer

(89 mM Tris, 89 mM borate and 2 mM EDTA). The major band was excised and eluted by crushing and soaking for 1216 h at 4 C with rotation in TEK (10 mM Tris-HCl, pH 7.58.0, 1 mM EDTA and 250 mM KCl) for 7GGT and TESr (10 mM Tris-HCl, pH 7.58.0, 1 mM EDTA and 250 mM SrCl2) for 22GGG, and ethanol precipitated for 216 h at 20 C. The precipitated product was resuspended in

10 mM Tris-Cl, pH 7.5.

G-quadruplex formation and purication. 7GGT (1 mM) or 22GGG (700 mM) were heat denatured in K hTel buffer (50 mM Tris-HCl, pH 8.0, 1 mM MgCl2 and 150 mM KCl) or Sr2 hTel buffer (50 mM Tris-HCl, pH 8.0, 1 mM MgCl2 and2.5 mM SrCl2), respectively, for 5 min at 95 C. They were allowed to cool slowly (B1 h) to 25 C and left to equilibrate at this temperature for 72 h. The folded DNA was added to 6 native gel loading buffer (0.25% bromophenol blue, 0.25%

xylene cyanol and 30% glycerol). 7GGT was electrophoresed on a non-denaturing 12% polyacrylamide gel containing 150 mM KCl for 4.5 h at 12 W at 22 C. 22GGG was electrophoresed on a non-denaturing 12% polyacrylamide gel containing2.5 mM SrCl2 for 24 h at 40 V at 18 C. Both buffer and gel contained the same constituents as the DNA-folding buffer. Ultraviolet shadowing was used to conrm the location of the DNA bands on the gel. The band of interest was excised and crushed in either TEK (7GGT) or TESr (22GGG) and incubated for 216 h at 4 C with rotation. The supernatant was ltered (0.22 mm) and the DNA precipitated with ethanol for 216 h at 20 C. The precipitated product was resuspended in

the original folding buffer. DNA concentrations were determined by ultraviolet absorbance at 260 nm (7GGT: 69,800 M 1 cm 1; 22GGG: 228,500 M 1 cm 1;

9C: 76,500 M 1 cm 1). Concentrations of G-quadruplexes are given as the concentration of assembled complexes (that is, taking strand stoichiometry into account). Folded G-quadruplexes were stored at 4 C until use.

In some experiments, oligonucleotides were 50-end-labelled with [g-32P]ATP prior to G-quadruplex formation and purication, as described23. Radiolabelled structures were used for some native gel analyses (for example, Fig. 1a), and P1 digestion assays, at 1015 103 c.p.m. per gel lane, whereas unlabelled

G-quadruplexes were used for telomerase assays, CD analysis and other native gel analyses (for example, Fig. 1c). In the latter experiment, 250 ng of each DNA was electrophoresed on a non-denaturing 12% polyacrylamide gel, which was stained in 1 SYBR Gold (Life Technologies) at 25 C for 30 min, and visualized on a

Typhoon FLA9500 scanner (GE Healthcare Lifesciences) using a 488-nm laser and a 526 BP emission lter.

Circular dichroism. CD spectra were recorded at 25 C on a Jasco J-810 spectrometer or an Aviv 215S CD spectrometer equipped with a Peltier temperature controller. G-quadruplex samples of the desired conformation were prepared at 20 mM in corresponding hTel buffers. Four scans were accumulated over the wavelength range 220320 nm in a 0.1-cm pathlength cell at standard sensitivity, data pitch 0.1 nm, continuous scanning mode, scanning speed 100 nm min 1, response 4 s and bandwidth 1 nm. Buffers alone were also scanned and these spectra subtracted from the average scans for each sample. CD spectra were

collected in units of millidegrees, normalized to the total species concentrations and expressed as molar ellipticity units (deg cm2 dmol 1).

Electrospray ionization mass spectrometry. All ESI mass spectra were obtained using a Waters Q-TOF Ultima ESI mass spectrometer (Manchester, UK). Puried G-quadruplex samples were dissolved in 150 mM NH4OAc at a concentrationof 20 mM. In all experiments the capillary voltage (2.2 kV), cone voltage (35 V), desolvation temperature (40 C), radio frequency lens (65 V), desolvation gas ow (150 l h 1), collision energy (4 V), cone gas ow (0.0021 l h 1) and TOF (9.1 kV)

remained the same. All spectra were obtained in the negative ion mode. The instrument was calibrated using 1 mg ml 1 caesium iodide. Samples were injected at a ow rate of 10 ml min 1 with a Harvard Model 22 syringe pump (Natick,

USA).

Complementary C-strand trap assay. This method was carried out using a modied procedure to that published37,38. 32P-labelled gel-puried [7GGT]4 (3,000 c.p.m. ml 1, 4 mM) or unlabelled [22GGG]2 (2.5 mM) were incubated in the presence of a 10-fold excess of 9C (Table 1) at 30 C to give a nal reaction volume of 28 ml. Aliquots (4 ml) of this hybridization reaction were removed at regular time intervals and loaded onto a native 12% polyacrylamide gel containing 150 mM KCl ([7GGT]4) or a native 16% polyacrylamide gel containing 2.5 mM

SrCl2 ([22GGG]2), with the gel running continuously between time points. Electrophoresis was conducted as described above. For complementary C-strand trap assays with increasing 9C concentrations, 32P-labelled gel-puried [7GGT]4 was incubated at 11 mM in the presence of a 5-, 10-, 25- or 100-fold excess of 9C at either 30 C or 37 C. For [7GGT]4, the gel was dried at 80 C for 40 min, exposed to a PhosphorImager screen, visualized on a Typhoon FLA9500 scanner and analysed using ImageQuant software. For [22GGG]2, the gel was stained with

SYBR Gold at 25 C for 30 min, visualized on a Typhoon FLA9500 scanner and analysed using ImageQuant software. Uncropped versions of blots are provided in Supplementary Fig. 7.

Phosphodiesterase 1 digestion of G-quadruplexes. 32P-end-labelled [7GGT]4 or [22GGG]2 (15,000 c.p.m.) and their linear controls were incubated with 2 mg P1 (Affymetrix; resuspended in 110 mM Tris-HCl, pH 8.9, 110 mM NaCl, 15 mM MgCl2 and 50% glycerol) in hTel buffer (as listed above) for 1 h at 30 C in a total reaction volume of 20 ml. The reaction was terminated with 80 ml of stop-buffer (50 mM Tris-HCl, pH 8.3, 20 mM EDTA, 0.2% SDS and 12 103 c.p.m. of a

50-32P-labelled synthetic 100-mer DNA as an internal recovery standard). From this point, the samples were treated in the same manner as telomerase activity assay reactions (see below).

HEK293T fermentation and telomerase overexpression. HEK293T cells (from Dr Timothy Adams, Commonwealth Scientic and Industrial Research Organisation) were adapted to grow in suspension in Freestyle 293 Expression medium (Life Technologies) supplemented with 200 mg l 1 G418 (Life

Technologies) using a humidied shaker incubator (37 C, 5% CO2, 130 r.p.m.). The adapted HEK293T culture was maintained in Erlenmeyer shaker asks and scaled up in a 20-l WAVE bioreactor (GE Healthcare), seeded at an initial working volume of 5 l at a concentration of 0.8 106 viable cells per ml. To reduce shear

stress, Pluronic F86 (Life Technologies) was added to the culture at 0.2% w/v nal concentration. The culture was scaled up to 20 l at a viable cell density of 3 106

cells per ml, at which time transient transfection was initiated. The hTERT gene54 under a CMV promoter was cloned into plasmid pAPEX-3P40,55. The hTR gene under a U3 promoter56 and dyskerin (OriGene Technologies) under a CMV promoter were cloned into a single plasmid in vector pAPEX-3 (ref. 55). Plasmids were produced on the 200-mg scale by GenScript (USA) and used at an hTERT:hTR ratio of 1:19 by weight. Polyethylenimine (25 kDa linear, Polysciences) was prepared as an aqueous solution at 1 mg ml 1 at pH 7.0 and lter-sterilized(0.22 mm). Per litre of culture, a transfection mix was prepared by adding 1 mg DNA to 100 ml PBS pre-warmed to 37 C, followed by addition of 4 ml polyethylenimine solution; the solution was mixed gently and incubated at room temperature (RT) for 15 min prior to addition to the 20-l culture, which was maintained at 37 C with a rocking speed of 25 r.p.m. and rocking angle of 9. At 2 days after transfection, the culture was fed with 5 g l 1 Lupin (Cell

Biosciences), 2 mM Glutamax-1 (Life Technologies) and 5 g l 1 glucose (Sigma), and the temperature was reduced to 32 C. The cells were harvested 4 days after transfection by centrifugation (1,500g, 10 min, 4 C), snap-frozen on liquid nitrogen and stored at 80 C. Telomerase high-expressing HEK293T cell pellets

are available from Abbexa Ltd, Cambridge, UK.

Purication of overexpressed 293T telomerase. Approximately 130 g 293T cell mass (B4 1010 cells) was broken up to a paste in a 1-l plastic beaker. The cell

paste was suspended in 800 ml ice-cold lysis buffer (10 mM HEPES-KOH (pH 8.0), 20 mM KCl, 2 mM MgCl2, 1% v/v Triton X-100, 1 mM dithiothreitol (DTT)). Once suspended, 8 ml of phenylmethylsulfonyl uoride solution (100 mM in ethanol, made fresh) was added with stirring. The suspension was transferred in 50-ml portions to a 50-ml Dounce homogenizer with a tight pestal on ice and processed

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 9

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

to ensure complete dissolution of cell paste, then pooled into a 1-l bottle equipped with magnetic stir bar. The lysate was stirred on ice for 1 h, then divided between four 250-ml bottles for the Beckman JA-14 rotor. The lysate was claried with centrifugation at 14,000 r.p.m. (B30,000g) at 2 C for 30 min. The clear lysate (B850 ml) was collected into a 1-l bottle equipped with magnetic stirrer. With stirring on ice, 22 ml of 2 M MgCl2 (B50 mM nal Mg2 ) was added and the suspension was stirred for 1 h to selectively precipitate ribonucleoprotein complexes57. The suspension was divided between four 250-ml bottles for the Beckman JA-14 rotor, and the products were collected with centrifugation at 10,000 r.p.m. (B15,000g) at 2 C for 30 min. The supernatant was decanted to provide white pellets. Each pellet was suspended in 50 ml immunoprecipitation (IP) buffer(50 mM HEPES-KOH (pH 8.0), 500 mM KCl, 2 mM MgCl2, 1% v/v Triton X-100, 10% v/v glycerol and 1 mM DTT). To aid dissolution, the suspension was processed through a Dounce homogenizer with a tight pestal on ice until a slightly white, homogeneous solution was obtained. The material was pooled into a 500-ml bottle on ice, and the volume was made up to 300 ml with additional IP buffer. Polyclonal hTERT antibody (raised against hTERT amino acids 276294 (ARPAEEATSLEGALSGTRH)41; available from Abbexa Ltd) (12 mg 40 mg ml 1

nal concentration) was added and the solution shaken on ice for 30 min. To capture the antibodytelomerase complex, 12 ml of a 50% v/v slurry of Protein G/sepharose (GE Healthcare) was added (20 ml beads per ml nal concentration)

and the suspension was shaken on ice for 1.5 h. Working in a cold room, the immunoprecipitate was collected into a 50-mm diameter fritted glass column (Bio-Rad) with vacuum suction and then washed with 200 ml ice-cold IP buffer. The immunoprecipitate was suspended in 20 ml telomerase buffer (50 mM HEPESKOH (pH 8.0), 300 mM KCl, 2 mM MgCl2, 0.1% v/v Triton X-100, 10% v/v glycerol and 1 mM DTT) containing 3 mg peptide ARPAEEATSLEGALSGTRH (20 molar equiv per antibody, available from Abbexa Ltd). The suspension was incubated at RT for 1 h with gentle shaking every 5 min, then the eluate was either collected (for use in assays requiring immunopuried telomerase) or transferred directly to a 25-mm diameter fritted glass column (Bio-Rad) containing 15 nmol gel-puried 50-BIOTIN-CTAGACCTGTCATCA(TTAGGG)3-30 oligonucleotide immobilized on 500 ml UltraLink Neutravidin Plus beads (Thermo-Fisher) (for the highly puried preparations in Figs 3d and 4b). The suspension was rotated at RT for 30 min and then at 4 C for 1 h. For the work in this paper, it was necessary to elute telomerase in the presence of two different solution-phase DNA traps,so that this DNA (traces of which remain in the puried telomerase preparation) was identical to the substrates being tested: 6 ml of suspension was collected in each of two 0.5-ml micro-spin columns (GE Healthcare), and washed with 3 ml cold telomerase buffer. After washing, columns were centrifuged for 10 s at 2,400 g to

remove residual buffer. Each bead sample was suspended in 500 ml telomerase buffer, followed by addition of 25 nmol of one of two different solution-phase DNAs dissolved in 100 ml telomerase buffer: (i) 7GGT (Table 1) or (ii) [7GGT]4

G-quadruplex. Activity-dependent elution was initiated by addition of 6 mldTTP dATP (10 mM each deoxyribonucleotide triphosphate, nal concentration

0.1 mM). The suspensions were rotated at RT for 30 min and the product solutions collected with centrifugation at 2,400g for 10 s. Each product solution was layered on an 11-ml 10-40% glycerol gradient composed of (20 mM HEPES-KOH(pH 8.0), 300 mM KCl, 2 mM MgCl2, 0.1% w/v octyl b-D-glucopyranoside and1 mM DTT) in Beckman Ultra-clear centrifuge tubes (14 89 mm) for the

Beckman SW-41 rotor. Telomerase was sedimented with centrifugation at 35,000 r.p.m. (B210,000g at rmax) at 4 C for 20 h. The tube was punctured at the bottom with a 30-gauge needle, and 0.5-ml fractions were collected by gravity. Fractions were assayed for telomerase concentration by dot-blot northern against hTR as described40, and equal amounts of enzyme used in each assay. Telomerase typically eluted in fractions 79; B35 pmol telomerase was obtained from each gradient. For extension assays using linear 7GGT or [7GGT]4 as substrates(Fig. 3d), the enzyme prepared in the presence of the equivalent trap DNA was used (with the [7GGT]4-trapped enzyme being used for the No DNA lanes).

Enzyme trapped with linear 7GGT was dialysed into telomerase buffer containing2.5 mM SrCl2, and equal amounts of this enzyme used for the extension of 22GGG and [22GGG]2 in Fig. 4b.

SDSPAGE and silver-staining analysis of puried telomerase. A 50-ml aliquot of puried telomerase, eluted in the presence of [7GGT]4 G-quadruplex DNA, was placed in a Pierce Slide-a-Lyzer mini-dialysis cup (MWCO 3,500) and dialysed at RT for 30 min against 200 ml of (20 mM HEPES-KOH (pH 8.0), 20 mM KCl, 2 mM MgCl2, 20% v/v glycerol, 0.1% w/v octyl b-D-glucopyranoside and 1 mM DTT).

For SDSPAGE, 30 ml of the dialysed solution was combined with 10 ml 10% w/v SDS, 5 ml 1 M DTT and 5 ml 4 NuPAGE LDS loading buffer (Life Technologies).

The sample was denatured at 80 C for 10 min, cooled to RT and centrifuged at 16,000g for 1 min. A 20-ml aliquot was electrophoresed over a NuPAGE 412% bis-Tris gradient mini-gel at 100 V for 3 h. For the molecular weight marker, the Life Technologies Benchmark protein ladder was diluted 100-fold in 1 NuPAGE

LDS loading buffer; from this, 3 ml was diluted with 20 ml 1 LDS buffer and

loaded on the gel. Silver staining was performed with the Life Technologies SilverXpress kit.

DHX36 western blotting. A 25-ml aliquot of immunopuried telomerase or four-step puried telomerase, eluted in the presence of [7GGT]4 G-quadruplex

DNA, was combined with: 20 ml H2O, 25 ml 4 NuPAGE LDS loading buffer, 10 ml

1 M DTT and 20 ml 10% w/v SDS. The samples were denatured at 80 C for 10 min, cooled to RT and centrifuged at 16,000g for 1 min. A 30-ml aliquot of each was electrophoresed over a NuPAGE 412% bis-Tris gradient mini-gel at 100 V for 4 h. For the molecular weight marker, 2 ml of the Life Technologies MagicMark-XP was used. Proteins were transferred to polyvinylidene diuoride membranes at 30 V for 90 min. Blocking, washing, secondary anti-rabbit and chemiluminescent detection were performed using the Western Breeze reagents from Life Technologies. The primary anti-DHX36 antibody was polyclonal from rabbit, Abcam #70269, diluted 1:500 in primary antibody diluent; probing was performed at RT for 2 h. Figure 3c represents a 10-min exposure.

Dyskerin western blotting. HEK293T cell pellets were resuspended in4 NuPAGE LDS loading buffer, 2% b-mercaptoethanol and 2% Benzonase

nuclease (Merck Millipore) at 10,000 cells per ml. The samples were denatured at 68 C for 10 min and centrifuged at 16,000g for 1 min. A 5-ml aliquot of each was electrophoresed over a NuPAGE 412% bis-Tris gradient mini-gel at 100 V for 2 h. For the molecular weight marker, 5 ml of the Bio-Rad Preision Plus prestained marker was used. Proteins were transferred to nitrocellulose membrane at 100 V for 60 min. Membranes were blocked using 5% skim milk, probed with primary antibodies diluted in 1% skim milk (mouse monoclonal anti-vinculin 1:5,000 (Sigma #V9131) or polyclonal rabbit anti-dyskerin (in-house) 1:1,000) for 1 h at RT and washed 5 5 min in TBST (0.24% Tris-Cl, 0.05% Tris, 0.8% NaCl and 0.1%

Tween-20, pH 7.6). Membranes were then probed with horseradish peroxidaseconjugated secondary antibodies (goat anti-mouse or goat anti-rabbit; DAKO) diluted 1:5,000 in 1% milk. Detection was performed with Amersham ECL Prime western blotting detection reagent and a FujiFilm Las4000 luminescent image analyzer, with autoexposure.

Telomerase activity assays. The following reaction was prepared to give 15 ml per sample: between 1 nM and 2 mM of the specied oligonucleotide, 50 mM Tris-HCl pH 8.5, 1 mM MgCl2, 5 mM DTT, 1 mM spermidine-HCl, 0.5 mM dTTP, 0.5 mM dATP, 4.6 mM non-radioactive dGTP and 0.33 mM [a-32P]dGTP at 20 mCi ml 1, 6,000 Ci mmol 1 (PerkinElmer Life Sciences), 10% glycerol and either 150 mM

KCl for experiments with 7GGT or 2.5 mM SrCl2 for experiments with 22GGG. The reaction was initiated by adding 5 ml of puried human telomerase, and incubating at 30 C for 1 h. The reaction was quenched by the addition of 80 ml of stop-buffer (50 mM Tris-HCl, pH 8.3, 20 mM EDTA and 0.2% SDS) and 12 103 c.p.m. of a 50-32P-labelled synthetic 100-mer DNA as an internal

recovery standard. The solution was extracted with an equal volume of phenol/ chloroform/isoamyl alcohol (25:24:1, v/v/v) and precipitated with ethanol in the presence of 2.5 M NH4OAc. The pellet was air-dried at RT for 15 min and dissolved in 5 ml TE buffer (10 mM Tris-HCl, pH 8.0 and 1 mM EDTA, pH 8.0), followed by addition of 5 ml formamide buffer (90% deionized formamide, 0.1% bromophenol blue and 0.1% xylene cyanol in 1 TBE). The solution was heated at 90 C for

5 min, and 3 ml was electrophoresed over a 10% polyacrylamide sequencing gel(0.2 mm thick 40 cm length 35 cm width, 32-well comb) run in 1 TBE/8 M

urea at 85 W. The gel was transferred to lter paper, dried for 30 min at 80 C, exposed to a PhosphorImager screen, visualized on a Typhoon FLA9500 scanner (GE Healthcare Lifesciences) and analysed using ImageQuant software. The total intensities of extension products were normalized against the intensity of the

32P-labelled 100-mer recovery and loading control. Resulting values were expressed as a percentage of the reaction with maximal activity and plotted against substrate concentrations and tted to the MichaelisMenten equation to give Km values.

For experiments where dTTP was titrated, the oligonucleotide concentration was kept at 1 mM and the concentration of dTTP was changed as outlined in the gure legend. All other reagents and procedures remained the same. Activity assays testing templated addition included 12 mM of the specied oligonucleotide and the following combinations and concentrations of labelled and unlabelled nucleotides; all other reaction conditions remained the same.

7GGT: T,A lanes: 32P-dTTP (25 mM, 80 Ci mmol 1) and dATP (0.5 mM). T,A,ddG lanes: 32P-dTTP (25 mM, 80 Ci mmol 1), dATP (0.5 mM) and ddGTP(0.5 mM).

T,A,G lanes: 32P-dTTP (25 mM, 80 Ci mmol 1), dATP (0.5 mM) and

dGTP (0.5 mM) or dTTP (0.5 mM), dATP (0.5 mM) and 32P-dGTP (5 mM, 400 Ci mmol 1).

22GGG:T,A lanes: 32P-dTTP (3 mM, 150 Ci mmol 1) and ddATP (25 mM). T,A,ddG lanes: 32P-dTTP (3 mM, 160 Ci mmol 1), dATP (0.5 mM) and ddGTP (25 mM).

T,A,G lanes: dTTP (0.5 mM), dATP (0.5 mM) and 32P-dGTP (5 mM, 400 Ci mmol 1).

Experiments in which reaction products were subjected to native gel electrophoresis contained 2 mM DNA, 0.7 mM non-radioactive dTTP and 0.33 mM [a-32P]dTTP at 10 mCi ml 1, 3,000 Ci mmol 1, with all other reaction conditions remaining the same. The reaction was quenched by the addition of 20 ml of stop-buffer (see above) and 8 103 c.p.m. of a 50-32P-labelled synthetic 100-mer

DNA as an internal recovery standard. The product was puried of unincorporated deoxyribonucleotide triphosphates using two rounds of mini Quick Spin Oligo column purication (Roche). Prior to use, Quick Spin Oligo columns were equilibrated by centrifugation twice at 1,000g for 10 min with 400 ml Sr2 hTel

10 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643 ARTICLE

buffer (see above). The puried product was added to 6 native gel loading buffer

(see above) and electrophoresed on a non-denaturing 16% polyacrylamide gel containing 2.5 mM SrCl2 for B48 h at 40 V at 18 C. Both buffer and gel contained the same constituents as the DNA-folding buffer. The gel was transferred to lter paper, dried for 30 min at 80 C, exposed to a PhosphorImager screen, visualized on a Typhoon FLA9500 scanner and analysed using ImageQuant software.

Expression and purication of G-quadruplex antibody (BG4). BG4-encoding plasmid (from the laboratory of Professor Shankar Balasubramanian, University of Cambridge, UK) was transformed into BL21(DE3) competent cells (Stratagene), which were cultured in TY media (1.6% tryptone peptone, 1% yeast extract and0.5% NaCl) and 50 mg ml 1 kanamycin. BG4 antibody expression was induced with 0.5 mM isopropyl b-D-1-thiogalactopyranoside for 3 h at 37 C. The cells were lysed in TES buffer (50 mM Tris-Cl pH 8.0, 1 mM EDTA and 20% sucrose) on ice for 10 min, diluted twofold in water, and left on ice for a further 10 min prior to centrifugation at 10,000g at 4 C for 30 min. The supernatant was ltered (0.2 mm)

and puried on a HIS-Select Nickel Afnity column (Sigma). The column was washed in 10 mM imidazole pH 8.0 in PBS, and BG4 antibody eluted in 250 mM imidazole pH 8.0 in PBS. BG4 antibody was concentrated and buffer exchanged into PBS in an Amicon Ultra-15 Centrifugal Filter Unit (Millipore). The concentration of BG4 was determined using Thermo Scientic Pierce BCA Protein Assay kit, and the antibody was stored at 20 C.

Cell synchronization and short interfering RNA (siRNA) transfection.

HEK293T mid-S-phase cell synchronization was performed using cells released from a thymidine/aphidicolin block as previously described49. Dyskerin knockdown was performed by reverse transfecting 120 pmol of Invitrogen custom-designed Stealth siRNA targeting either the dyskerin 30 untranslated region (siDKC#1, 50-AAGGCCACUUGAAGCUGGAGGAGAA-30) or the coding region (siDKC#2, 50-GGCCAAGATTATGCTTCCAGGTGTT-30). Cells were transfected using Life Technologies Lipofectamine RNAiMAX Transfection Reagent. Qiagen All Stars negative-control siRNA was used as a siRNA control.

Immunouorescence and telomere/telomerase FISH. Immunouorescence and FISH for telomeres and telomerase RNA (hTR) were performed as described49, with the following modications. Cytoplasm was removed by incubating slides in cytoplasmic extraction buffer (20 mM HEPES-KOH, pH 7.9, 20 mM NaCl, 5 mM MgCl2, 300 mM sucrose and 0.5% (v/v) NP-40) for 10 min with no shaking. Slides were washed once with PBS 0.1% Tween-20, once with PBS and then xed in 2%

paraformaldehyde in PBS for 20 min. Cells were washed with PBS then incubated in 0.1% Triton X-100 in PBS for 10 min, followed by rinsing twice in PBS. Slides were ethanol dehydrated with 70% (2 min), 90% (2 min) and 100% ethanol (2 min) followed by air drying prior to overlaying with 30 ml of FISH buffer containing 5 ng each of ve Alexa Fluor 488-labelled anti-hTR oligonucleotides and 5 ng of Texas Red-labelled telomere probe49. Slides were heated to 80 C for 3 min and incubated in a humidied chamber overnight at 37 C and washed49. Slides were rexed in 2% paraformaldehyde in PBS, washed as before and overlaid with phosphate-buffered gelatin49 for 1 h. Slides were incubated with BG4 primary antibody(600 nM) for 1 h at 37 C, washed four times for 5 min with PBS 0.01% Tween-20

and then overlaid with secondary antibody (Rabbit anti-DYKDDDDK Tag Antibody, Cell Signalling; 1:800 dilution) for 1 h at 37 C. Cells were washed as before and a uorescently labelled tertiary antibody (Alexa Fluor 647 Donkey Anti-Rabbit IgG (H L) Antibody, Life Technologies; 1:1,000 dilution) was

overlaid at 37 C for 30 min, washed as before and counterstained with 40,6-diamidino-2-phenylindole. Staining was visualized at RT on a Zeiss Axio Imager M1 microscope, with a Plan-Apochromat 63 oil objective (numerical

aperture, 1.4), and an AxioCam MR digital camera (Carl Zeiss) with consistent exposure times between experiments. For presentation purposes, pixel intensity histograms were adjusted in Axiovision (Carl Zeiss), equally across all gure panels, and images were cropped in Adobe Photoshop.

References

1. de Lange, T. et al. Structure and variability of human chromosome ends. Mol. Cell. Biol. 10, 518527 (1990).

2. Moyzis, R. K. et al. A highly conserved repetitive DNA sequence, (TTAGGG)n,

present at the telomeres of human chromosomes. Proc. Natl Acad. Sci. USA 85, 66226626 (1988).3. Harley, C. B., Futcher, A. B. & Greider, C. W. Telomeres shorten during ageing of human broblasts. Nature 345, 458460 (1990).

4. Allsopp, R. C. et al. Telomere length predicts replicative capacity of human broblasts. Proc. Natl Acad. Sci. USA 89, 1011410118 (1992).

5. Greider, C. W. & Blackburn, E. H. Identication of a specic telomere terminal transferase activity in Tetrahymena extracts. Cell 43, 405413 (1985).

6. Greider, C. W. & Blackburn, E. H. The telomere terminal transferase of Tetrahymena is a ribonucleoprotein enzyme with two kinds of primer specicity. Cell 51, 887898 (1987).

7. Lingner, J. et al. Reverse transcriptase motifs in the catalytic subunit of telomerase. Science 276, 561567 (1997).

8. Feldser, D. M. & Greider, C. W. Short telomeres limit tumor progression in vivo by inducing senescence. Cancer Cell 11, 461469 (2007).

9. Shay, J. W. & Bacchetti, S. A survey of telomerase activity in human cancer. Eur. J. Cancer 33, 787791 (1997).

10. Williamson, J. R., Raghuraman, M. K. & Cech, T. R. Monovalent cation-induced structure of telomeric DNA: the G-quartet model. Cell 59, 871880 (1989).

11. Burge, S., Parkinson, G. N., Hazel, P., Todd, A. K. & Neidle, S. Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res. 34, 54025415 (2006).

12. Huppert, J. L. & Balasubramanian, S. Prevalence of quadruplexes in the human genome. Nucleic Acids Res. 33, 29082916 (2005).

13. Siddiqui-Jain, A., Grand, C. L., Bearss, D. J. & Hurley, L. H. Direct evidence for a G-quadruplex in a promoter region and its targeting with a small molecule to repress c-MYC transcription. Proc. Natl Acad. Sci. USA 99, 1159311598 (2002).

14. Kumari, S., Bugaut, A., Huppert, J. L. & Balasubramanian, S. An RNA G-quadruplex in the 5 UTR of the NRAS proto-oncogene modulates translation. Nat. Chem. Biol. 3, 218221 (2007).

15. Gomez, D. et al. Telomerase downregulation induced by the G-quadruplex ligand 12459 in A549 cells is mediated by hTERT RNA alternative splicing. Nucleic Acids Res. 32, 371379 (2004).

16. Smith, J. S. et al. Rudimentary G-quadruplex-based telomere capping in Saccharomyces cerevisiae. Nat. Struct. Mol. Biol. 18, 478485 (2011).

17. Schaftzel, C. et al. In vitro generated antibodies specic for telomeric guanine-quadruplex DNA react with Stylonychia lemnae macronuclei. Proc. Natl Acad. Sci. USA 98, 85728577 (2001).

18. Bif, G., Tannahill, D., McCafferty, J. & Balasubramanian, S. Quantitative visualization of DNA G-quadruplex structures in human cells. Nat. Chem. 5, 182186 (2013).

19. Shi, D. F., Wheelhouse, R. T., Sun, D. & Hurley, L. H. Quadruplex-interactive agents as telomerase inhibitors: synthesis of porphyrins and structure-activity relationship for the inhibition of telomerase. J. Med. Chem. 44, 45094523 (2001).

20. Cuesta, J., Read, M. A. & Neidle, S. The design of G-quadruplex ligands as telomerase inhibitors. Mini Rev. Med. Chem. 3, 1121 (2003).

21. De Cian, A. et al. Targeting telomeres and telomerase. Biochimie 90, 131155 (2008).

22. Zahler, A. M., Williamson, J. R., Cech, T. R. & Prescott, D. M.

Inhibition of telomerase by G-quartet DNA structures. Nature 350, 718720 (1991).

23. Oganesian, L., Moon, I. K., Bryan, T. M. & Jarstfer, M. B. Extension of G-quadruplex DNA by ciliate telomerase. EMBO J. 25, 11481159 (2006).

24. Brown, A. F. et al. A self-regulating template in human telomerase. Proc. Natl Acad. Sci. USA 111, 1131111316 (2014).

25. Blackburn, E. H. & Collins, K. Telomerase: an RNP enzyme synthesizes DNA. Cold Spring Harb. Perspect. Biol. 3, a003558 (2011).

26. Wang, H. & Blackburn, E. H. De nova telomere addition by Tetrahymena telomerase in vitro. EMBO J. 16, 866879 (1997).

27. Sauerwald, A. et al. Structure of active dimeric human telomerase. Nat. Struct. Mol. Biol. 20, 454460 (2013).

28. Jiang, J. et al. The architecture of Tetrahymena telomerase holoenzyme. Nature 496, 187192 (2013).

29. Mergny, J. L., De Cian, A., Ghelab, A., Sacca, B. & Lacroix, L. Kinetics of tetramolecular quadruplexes. Nucleic Acids Res. 33, 8194 (2005).

30. Pedroso, I. M., Duarte, L. F., Yanez, G., Baker, A. M. & Fletcher, T. M. Induction of parallel human telomeric G-quadruplex structures by Sr(2 ).

Biochem. Biophys. Res. Commun. 358, 298303 (2007).31. Chen, F. M. Sr2 facilitates intermolecular G-quadruplex formation of

telomeric sequences. Biochemistry 31, 37693776 (1992).32. Wang, Y. & Patel, D. J. Guanine residues in d(T2AG3) and d(T2G4)form parallel-stranded potassium cation stabilized G-quadruplexeswith anti glycosidic torsion angles in solution. Biochemistry 31, 81128119 (1992).

33. Williamson, J. R. G-quartet structures in telomeric DNA. Annu. Rev. Biophys. Biomol. Struct. 23, 703730 (1994).

34. Mergny, J. L., Li, J., Lacroix, L., Amrane, S. & Chaires, J. B. Thermal difference spectra: a specic signature for nucleic acid structures. Nucleic Acids Res. 33, e138 (2005).

35. Gornall, K. C., Samosorn, S., Talib, J., Bremner, J. B. & Beck, J. L. Selectivity of an indolyl berberine derivative for tetrameric G-quadruplex DNA. Rapid Commun. Mass Spectrom. 21, 17591766 (2007).

36. Rosu, F., Gabelica, V., Houssier, C., Colson, P. & Pauw, E. D. Triplex and quadruplex DNA structures studied by electrospray mass spectrometry. Rapid Commun. Mass Spectrom. 16, 17291736 (2002).

37. Oganesian, L., Graham, M. E., Robinson, P. J. & Bryan, T. M. Telomerase recognizes G-quadruplex and linear DNA as distinct substrates. Biochemistry 46, 1127911290 (2007).

NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 11

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8643

38. Raghuraman, M. K. & Cech, T. R. Effect of monovalent cation-induced telomeric DNA structure on the binding of Oxytricha telomeric protein. Nucleic Acids Res. 18, 45434552 (1990).

39. Zaug, A. J., Podell, E. R. & Cech, T. R. Human POT1 disrupts telomeric G-quadruplexes allowing telomerase extension in vitro. Proc. Natl Acad. Sci. USA 102, 1086410869 (2005).

40. Jurczyluk, J. et al. Direct involvement of the TEN domain at the active site of human telomerase. Nucleic Acids Res. 39, 17741788 (2011).

41. Cohen, S. B. & Reddel, R. R. A sensitive direct human telomerase activity assay. Nat. Methods 5, 355360 (2008).

42. Cohen, S. B. et al. Protein composition of catalytically active human telomerase from immortal cells. Science 315, 18501853 (2007).

43. Lattmann, S., Stadler, M. B., Vaughn, J. P., Akman, S. A. & Nagamine, Y. The DEAH-box RNA helicase RHAU binds an intramolecular RNA G-quadruplex in TERC and associates with telomerase holoenzyme. Nucleic Acids Res. 39, 93909404 (2011).

44. Sexton, A. N. & Collins, K. The 5 guanosine tracts of human telomerase RNA are recognized by the G-quadruplex binding domain of the RNA helicase DHX36 and function to increase RNA accumulation. Mol. Cell. Biol. 31, 736743 (2011).

45. Nicoludis, J. M., Barrett, S. P., Mergny, J. L. & Yatsunyk, L. A. Interaction of human telomeric DNA with N-methyl mesoporphyrin IX. Nucleic Acids Res. 40, 54325447 (2012).

46. Nicoludis, J. M. et al. Optimized end-stacking provides specicity of N-methyl mesoporphyrin IX for human telomeric G-quadruplex DNA. J. Am. Chem. Soc. 134, 2044620456 (2012).

47. Tippana, R., Xiao, W. & Myong, S. G-quadruplex conformation and dynamics are determined by loop length and sequence. Nucleic Acids Res. 42, 81068114 (2014).

48. Tomlinson, R. L. et al. Telomerase reverse transcriptase is required for the localization of telomerase RNA to Cajal bodies and telomerese in human cancer cells. Mol. Biol. Cell 19, 37933800 (2008).

49. Stern, J. L., Zyner, K. G., Pickett, H. A., Cohen, S. B. & Bryan, T. M. Telomerase recruitment requires both TCAB1 and Cajal bodies independently. Mol. Cell. Biol. 32, 23842395 (2012).

50. Ray, S. et al. RPA-mediated unfolding of systematically varying G-quadruplex structures. Biophys. J. 104, 22352245 (2013).

51. Zhang, M. L. et al. Yeast telomerase subunit Est1p has guanine quadruplex-promoting activity that is required for telomere elongation. Nat. Struct. Mol. Biol. 17, 202209 (2010).

52. Sen, D. & Gilbert, W. Formation of parallel four-stranded complexes by guanine-rich motifs in DNA and its implications for meiosis. Nature 334, 364366 (1988).

53. Canudas, S. et al. A role for heterochromatin protein 1g at human telomeres. Genes Dev. 25, 18071819 (2011).

54. Colgin, L. M. et al. The hTERTa splice variant is a dominant negative inhibitor of telomerase activity. Neoplasia 2, 426432 (2000).

55. Evans, M. J., Hartman, S. L., Wolff, D. W., Rollins, S. A. & Squinto, S. P. Rapid expression of an anti-human C5 chimeric Fab utilizing a vector that replicates in COS and 293 cells. J. Immunol. Methods 184, 123138 (1995).

56. Fu, D. & Collins, K. Distinct biogenesis pathways for human telomerase RNA and H/ACA small nucleolar RNAs. Mol. Cell 11, 13611372 (2003).

57. Palmiter, R. D. Magnesium precipitation of ribonucleoprotein complexes. Expedient techniques for the isolation of undergraded polysomes and messenger ribonucleic acid. Biochemistry 13, 36063615 (1974).

Acknowledgements

We are grateful to Timothy Adams (CSIRO, Parkville, VIC, Australia) for providing pAPEX vectors, and for consultation on overexpression of telomerase. We thank Jessie Zhang for help with siRNA transfections, and Grant Grifths and Nathan Zammit for contributing data at an early stage of the project. We thank Shankar Balasubramanian (University of Cambridge, Cambridge, United Kingdom) for providing the plasmid encoding the BG4 antibody. We thank the Australian Research Council for the mass spectrometers used in this work. This research was supported by project grants awarded by the Cancer Council NSW (RG11/07, RG12/01 and RG12/02) and the National Health and Medical Research Council (571073). K.G.Z. was supported by an Australian Postgraduate Award and a Cancer Institute NSW PhD Scholarship, A.L.M. by a Denise Higgins Scholarship, and T.M.B. by a Career Development Fellowship from the Cancer Institute NSW (11/CDF/3-05).

Author contributions

K.C.P., A.L.M. and T.M.B. conceived the project and designed experiments. A.L.M., K.C.P., S.B.C., K.G.Z., N.S. and T.M.B. carried out experiments. T.P. and G.O.L. performed large-scale culture of telomerase-overexpressing cells. K.C.P. and J.L.B. performed and interpreted mass spectrometry analysis. All authors contributed to writing the manuscript.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications .

Competing nancial interests: The hTERT antibody, CMRI 276-294, antigenic release peptide and telomerase overexpressing HEK293T cells are available to researchers from Abbexa Ltd, under exclusive license from the authors institution, Childrens Medical Research Institute.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Moye, A. L. et al. Telomeric G-quadruplexes are a substrate and site of localization for human telomerase. Nat. Commun. 6:7643 doi: 10.1038/ncomms8643 (2015).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

12 NATURE COMMUNICATIONS | 6:7643 | DOI: 10.1038/ncomms8643 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

Word count: 11777

Show less

Abstract

Translate

It has been hypothesized that G-quadruplexes can sequester the 3' end of the telomere and prevent it from being extended by telomerase. Here we purify and characterize stable, conformationally homogenous human telomeric G-quadruplexes, and demonstrate that human telomerase is able to extend parallel, intermolecular conformations in vitro. These G-quadruplexes align correctly with the RNA template of telomerase, demonstrating that at least partial G-quadruplex resolution is required. A highly purified preparation of human telomerase retains this extension ability, establishing that the core telomerase enzyme complex is sufficient for partial G-quadruplex resolution and extension. The parallel-specific G-quadruplex ligand N-methyl mesoporphyrin IX (NMM) causes an increase in telomeric G-quadruplexes, and we show that telomerase colocalizes with a subset of telomeric G-quadruplexes in vivo. The ability of telomerase to partially unwind, extend and localize to these structures implies that parallel telomeric G-quadruplexes may play an important biological role.

Details

Title

Telomeric G-quadruplexes are a substrate and site of localization for human telomerase

Author

Moye, Aaron L; Porter, Karina C; Cohen, Scott B; Phan, Tram; Zyner, Katherine G; Sasaki, Natsuki; Lovrecz, George O; Beck, Jennifer L; Bryan, Tracy M

Pages

7643

Publication year

2015

Publication date

Jul 2015

Publisher

Nature Publishing Group

e-ISSN

20411723

Source type

Scholarly Journal

Language of publication

English

DOI

https://doi.org/10.1038/ncomms8643

ProQuest document ID

1695020068

Telomeric G-quadruplexes are a substrate and site of localization for human telomerase

Jump to:

Full text

Abstract

Details

Suggested sources