Human chromosomes contain many guanine (G)-rich elements capable of forming four-stranded G-quadruplex (G4) structures (Bochman et al., 2012; Huppert and Balasubramanian, 2005; Maizels and Gray, 2013; Rhodes and Lipps, 2015). A planar G-quartet is formed when four Gs form hydrogen bonds in a cyclical manner via their Hoogsteen faces; such G-quartets stack on top of each other and form G-quadruplexes (Sen and Gilbert, 1988; Sundquist and Klug, 1989; Williamson et al., 1989). G-rich sequences are primarily located in promoter regions, intron and exon boundaries, origins of replication and telomeres (Bochman et al., 2012; Huppert and Balasubramanian, 2005; Maizels and Gray, 2013; Rhodes and Lipps, 2015). Vertebrate telomeres consist of many tandem repeats of the sequence TTAGGG, comprising a double-stranded region and a 3' G-rich overhang (Blackburn, 1991). Fluorescence microscopy studies have shown the existence of G4 structures at the telomeric regions of fixed human cells using G4-specific antibodies (Biffi et al., 2013; Liu et al., 2016). In vitro, telomeric DNA can adopt many different conformations of G4, including inter- or intramolecular forms, arranged in parallel or anti-parallel orientation depending upon the directionality of the DNA backbone (Burge et al., 2006). Conditions that may mimic those in the nucleus, including high DNA concentration and water depletion, favour parallel conformations of human telomeric G4 (Heddi and Phan, 2011; Miller et al., 2010; Renciuk et al., 2009). However, the in vivo conformation(s) and biological significance of telomeric G4 in human cells remain elusive and undetermined.

Telomere shortening occurs with every cell division in normal human somatic cells, since conventional DNA polymerases cannot completely synthesize the telomeric end (Harley et al., 1990; Olovnikov, 1971; Watson, 1972). Telomerase, a telomere-specific ribonucleoprotein enzyme complex, extends telomeric DNA in cancer cells, stem cells and cells of the germline, using a unique mechanism of processive rounds of reverse transcription (Greider and Blackburn, 1985; Hiyama and Hiyama, 2007; Kim et al., 1994; Morin, 1989; Podlevsky and Chen, 2012). Human telomerase minimally consists of the highly conserved telomerase reverse transcriptase protein (hTERT) and an RNA component containing an 11 nucleotide (nt) template sequence (hTR) (Wyatt et al., 2010). Telomerase extends its telomeric substrate by first binding to telomeric DNA, followed by hybridization of the RNA template with the DNA. Second, telomerase extends the substrate DNA by using its RNA as a template for nucleotide addition to the DNA 3' end. Once the template 5' boundary has been reached, telomerase translocates downstream, resulting in re-alignment of the RNA template with the new 3' end of the product DNA (Greider and Blackburn, 1989; Podlevsky and Chen, 2012).

One function of telomeric G4 may be to act as a ‘cap’, protecting the telomere from DNA degradation (Smith et al., 2011). It has been widely believed that G4 formation within the 3' telomeric overhang blocks telomere extension by telomerase, since early in vitro studies had shown that G4 structures inhibit telomerase extension of a telomeric DNA substrate (Zahler et al., 1991; Zaug et al., 2005). In addition, stabilization of G4 with small-molecule ligands has been shown to more effectively inhibit telomerase activity, suggesting that chemical stabilization of G4 structures may be a viable anti-cancer therapeutic strategy (De Cian et al., 2007a; Gomez et al., 2016; Sun et al., 1997). However, the above-mentioned studies did not distinguish between G4 conformations, and used oligonucleotides that likely folded into anti-parallel or ‘hybrid’ G4 forms as a telomerase substrate. A variety of helicases, such as RECQ5, WRN, BLM, FANC-J and RHAU, recognize and resolve G4 in a conformation-specific manner (Budhathoki et al., 2016; Liu et al., 2010; London et al., 2008; Tippana et al., 2016). Similarly, we have demonstrated that telomerase can bind and extend specific conformations of telomeric G4 - those that are parallel-stranded and intermolecular - and that this property of telomerase is well conserved from ciliates to human (Moye et al., 2015; Oganesian et al., 2007; Oganesian et al., 2006). This suggests that if telomeric G-quadruplexes were to adopt a parallel conformation in vivo, they may indeed be extended by telomerase, contrary to what had been previously hypothesized.

Telomerase extension reactions in the presence of individual nucleotides demonstrated that the 3' end of a G4 substrate aligns correctly with the RNA template of telomerase, suggesting that at least partial resolution of G4 structure is required for telomerase extension (Moye et al., 2015). However, the evidence for partial disruption of the G4 was indirect, and a mechanistic understanding of telomerase resolution of G4 DNA, in the absence of a known helicase function of hTERT, remained elusive. Furthermore, since all previously-tested substrates were both parallel-stranded and intermolecular, it was unknown which of these properties form the basis of telomerase recognition. To directly study the mechanistic details of parallel G4 unfolding and extension by human telomerase, here we combined ensemble telomerase enzymatic assays and single-molecule FRET (smFRET) measurements in vitro, using both intramolecular and intermolecular parallel G4 as FRET sensors. We provide direct evidence that wild-type human telomerase can unfold both intra- and intermolecular parallel G4 completely in the presence of dNTPs, using a 3-step mechanism. First, telomerase binds the G4, partially changing its conformation in a process that involves binding of the hTR template sequence. Second, telomerase adds individual nucleotides to the 3' end of the partially unfolded parallel G4. Lastly, the translocation of telomerase results in complete disruption of the G4 structure. Unexpectedly, stabilization of parallel intermolecular G4 using different G4 ligands did not inhibit telomerase unfolding and extension of this substrate. Overall, we provide a mechanistic explanation of conformation-specific telomerase extension of telomeric G4 at single-molecule resolution and demonstrate that small molecule-mediated inhibition of telomerase extension of telomeric G4 is topology-dependent.

Previously, we have shown that highly purified human telomerase can disrupt parallel intermolecular G4 structures and then extend them in a processive manner (Moye et al., 2015). Here, we provide the first mechanistic details of telomerase resolution of parallel G4 structures. Our data demonstrate that G4 unfolding occurs in three major steps: i) the telomerase template RNA invades the G-quartets in order to hybridize with the G-rich DNA, causing partial opening of the G4 structure; ii) telomerase extends the 3' end of this partially unwound structure, adding single nucleotides according to the RNA template, and iii) translocation of telomerase once nucleotide addition has reached the template boundary is likely to trigger complete unfolding of the G4 structure (Figure 10C). It has previously been difficult to distinguish whether it is the parallel or intermolecular nature of G4 structures that allows their recognition by telomerase. Here, we overcome this difficulty by exploiting the stabilization of intramolecular parallel G4 conferred by the modified nucleotide 2'F-araG (Abou Assi et al., 2017), enabling the demonstration that inter- and intramolecular parallel G4 are extended by telomerase using the same 3-step mechanism. These data reveal human telomerase to be a parallel G4 resolvase.

Several other proteins have been shown to assist telomerase extension of telomeric DNA by resolving antiparallel G4 in the substrate, including telomeric protein POT1 (Hwang et al., 2014; Zaug et al., 2005), a splice variant of hnRNP A2 (Wang et al., 2012), and the ciliate protein StyRecQL (Paeschke et al., 2008; Postberg et al., 2012). In contrast, we have verified that highly purified telomerase, free of contaminating helicases or other proteins, is able to extend [7GGT]₄ (Moye et al., 2015). Therefore, we conclude that telomerase itself has the ability to unwind parallel G4, without the assistance of interacting helicases.

Telomerase partially unfolds parallel G4, independently of telomerase catalytic activity, as indicated by the drop in FRET signal upon telomerase association with G4 (Figure 5A–E). The resulting stable intermediate FRET value is consistent with a partial separation of the FRET dyes in the absence of complete strand dissociation. We interpret these data to indicate that binding of telomerase RNA to the nucleotides at the 3' end of one DNA strand allows RNA-DNA hybridization to occur while resulting in a conformational perturbation of the 5' end of the G4, but the dyes are still sufficiently close together to give a non-zero FRET signal; molecular modeling supported this interpretation (Figure 10B and C). To investigate whether the template RNA of telomerase plays a critical role in this initial G4 unfolding, we carried out experiments in the presence of a short RNA oligonucleotide with a sequence complementary to the telomere repeat, mimicking the template region of hTR. This RNA molecule was able to partially unfold the structure in a similar manner to the whole enzyme, albeit with lower affinity. This observation suggests that it is the RNA component of telomerase, rather than the hTERT protein, that promotes the initial unwinding of parallel G4, followed by hybridization to the 3' end. A DNA oligonucleotide of identical sequence did not promote any G4 unwinding (Figure 5H–J); if oligonucleotide binding was due to passive trapping of a spontaneously-unfolding G4, this would also have been observed in the presence of a DNA C-strand. Previous studies with relatively unstable antiparallel telomeric G4s in Na⁺ have observed passive trapping of the unfolded G4 by a peptide nucleic acid (PNA) oligonucleotide; the unfolding was zero order with respect to the PNA molecule, suggesting that the initial step is a rate-determining internal rearrangement of the quadruplex (Green et al., 2003). In contrast, both parallel G4s used in our experiments were highly stable; neither demonstrated any spontaneous unwinding during the observation time of the smFRET experiments, and in our previous work we have shown that the parallel tetrameric G4 did not show any unwinding over the time scale of hours (Moye et al., 2015).Telomeric protein POT1 unwinds a G4 through a conformational selection mechanism in which G4 unwinding occurs prior to POT1 binding (Chaires et al., 2020); each of the OB folds of two POT1 molecules then binds to one telomeric repeat of the 4-repeat G4 in a stepwise manner (Hwang et al., 2012). For telomerase, we favour a model in which the enzyme binds the G4 while it is still structured, followed by invasion by the RNA template; confirmation of this model will require further kinetic analyses. A similar template-mimicking RNA was unable to form a DNA-RNA hybrid with an antiparallel G4 of similar stability to the parallel G4 used in the present study (Birrento et al., 2015), demonstrating that this is a specific interaction between this RNA and parallel G4 structures. Single-molecule FRET studies have illuminated the mechanisms used by other G4-unwinding proteins; for example, the helicase RHAU/DHX36 was shown to stack on the top of the G4 plane and bind to the exposed 3' single stranded DNA tail (Chen et al., 2018), whereas the replication protein RPA has been proposed to bind to G4 structures via their exposed loops (Ray et al., 2013). In contrast, our data demonstrate that telomerase possesses a unique mechanism of G4 resolution, mediated by its integral RNA subunit.

Telomerase extends the 3' end of parallel G4 while the G4 remains partially structured, as demonstrated by the intermediate FRET state remaining unchanged until the hTR template boundary was reached (Figure 6). Although these experiments involved incorporation of a chain terminating ddGTP, it has been demonstrated by smFRET that ddGTP incorporation is compatible with subsequent DNA realignment and rehybridization with the upstream region of the template (Parks and Stone, 2014); it is therefore likely that conformational changes involved in the translocation process result in complete G4 unfolding, represented by a second step-wise drop in FRET signal. One possible explanation for this observation is that the G4 structure needs to pass through the telomerase DNA-binding cleft during the conformation changes required for telomerase translocation, and the resulting steric hindrance results in complete dissociation of the parallel telomeric G4. Our molecular modeling supports an alternative scenario, in which an hTERT protein conformational change during translocation results in opening of the active site, allowing the strands of the G4 to move away from each other (Figure 10C). This would be consistent with a ‘pivoting’ of the CTE away from the catalytic center that has been proposed to allow room for DNA-RNA dissociation and movement during translocation (Wu et al., 2017; Yang and Lee, 2015). In the case of a G4 substrate, it is possible that slight pivoting of the CTE is needed to allow the structure to bind in the enzyme active site (Figure 10C, middle panels). After nucleotide addition, when the end of the template is reached, this may act as a trigger for full pivoting of the CTE to allow translocation to occur (Wu et al., 2017; Figure 10C, right panel; the model in Figure 10A also shows the fully-extended structure). This may allow full dissociation of the strands of the G4, resulting in a drop in FRET to its lowest level.

Quantitation of the rates of the transition from the first FRET drop to the second support the notion that telomerase-mediated G4 unfolding is a complex multistep process, consisting of at least five kinetic steps. This is consistent with the requirement for extension of the DNA to the template boundary and translocation, processes which involve many kinetic substeps: nucleotide binding (Tomlinson et al., 2015), polymerization at each template position (Tomlinson et al., 2015), DNA primer realignment with the upstream region of the RNA (Parks and Stone, 2014), and conformational rearrangements to reposition the DNA-RNA hybrid into the active site (Parks and Stone, 2014).

We used the G4-stabilizing ligands NMM, SST16 and PhenDC3, all of which bind to parallel G4 (Lacroix et al., 2011; Nicoludis et al., 2012a; Samosorn et al., 2009), to understand whether telomerase extension of these structures can be modulated. Despite substantial thermal stabilization of the G4, these ligands did not inhibit telomerase extension of intermolecular parallel telomeric G4, with one of them instead slightly accelerating G4 unfolding. The ligands did reduce the number of intramolecular parallel G4 molecules unwound by telomerase; nevertheless, some molecules were unwound by telomerase despite ligand binding, using the same two-step mechanism as in the absence of ligands. The fact that telomerase was able to extend parallel G4 even after substantial thermal stabilization of the structure using either small molecule ligands (Figures 7–9) or chemical modification (Figure 1) provides evidence that thermal stabilization does not provide a barrier to telomerase-mediated unwinding.

NMM has been demonstrated to bind to parallel telomeric G4 by stacking on the outermost G-quartets (Nicoludis et al., 2012b), which would be predicted to interfere with telomerase extension. This observation suggests that, like other proteins such as RHAU helicase (Gueddouda et al., 2017; Tippana et al., 2016), telomerase is able to dislodge ligands from the tetramolecular G4 structure while unfolding it. This displacement activity is structure dependent, since telomerase is less able to dislodge ligands from a parallel intramolecular G4, possibly due to the presence of DNA loops. Such a displacement mechanism would have implications for interpreting the biological effects of G4-stabilizing ligands in cell-based studies, since any effects on telomerase action and telomere length will depend on the conformational specificity of the ligand.

There is growing evidence for G4 formation at telomeres (Biffi et al., 2013; Moye et al., 2015; Murat and Balasubramanian, 2014; Paeschke et al., 2005; Schaffitzel et al., 2001; Zhang et al., 2010). In human cells, it is not known if they form at the 3' end of the telomeric overhang, internally, or a combination of both, or which conformation of G4 occurs at these locations in vivo. While in-cell NMR studies of transfected telomeric oligonucleotides detected hybrid and antiparallel structures (Bao et al., 2019), a study using parallel-specific antibodies reported detection of parallel telomeric G-quadruplexes in human cells (Liu et al., 2016) and a parallel-specific ligand induces G4 formation at human telomeres (Moye et al., 2015). It is also possible that different conformations exist at different telomeric overhangs within the same cell. Intermolecular parallel G4 may also form between telomeric overhangs; it has been suggested that tetrameric parallel G-quadruplexes could be responsible for the correct alignment of four chromatids during meiosis (Sen and Gilbert, 1988). Regardless of this uncertainty, our data unequivocally demonstrate that human telomerase interacts directly with parallel G4 structures and resolves them. The conserved ability of human and ciliate telomerase enzymes to unwind these structures suggests that parallel G-quadruplexes form at telomeric overhangs in vivo, and do not form a barrier to telomerase extension.

All data generated or analysed during this study are included in the manuscript and supporting files. Source data files have been provided for all summary graphs.

1. 1. Abou Assi H
   2. El-Khoury R
   3. González C
   4. Damha MJ

   (2017) 2'-Fluoroarabinonucleic acid modification traps G-quadruplex and i-motif structures in human telomeric DNA

   Nucleic Acids Research 45:11535–11546.

   https://doi.org/10.1093/nar/gkx838

   • PubMed
   • Google Scholar
2. 1. Arthanari H
   2. Basu S
   3. Kawano TL
   4. Bolton PH

   (1998) Fluorescent dyes specific for quadruplex DNA

   Nucleic Acids Research 26:3724–3728.

   https://doi.org/10.1093/nar/26.16.3724

   • PubMed
   • Google Scholar
3. 1. Bao HL
   2. Liu HS
   3. Xu Y

   (2019) Hybrid-type and two-tetrad antiparallel telomere DNA G-quadruplex structures in living human cells

   Nucleic Acids Research 47:4940–4947.

   https://doi.org/10.1093/nar/gkz276

   • PubMed
   • Google Scholar
4. 1. Biffi G
   2. Tannahill D
   3. McCafferty J
   4. Balasubramanian S

   (2013) Quantitative visualization of DNA G-quadruplex structures in human cells

   Nature Chemistry 5:182–186.

   https://doi.org/10.1038/nchem.1548

   • PubMed
   • Google Scholar
5. 1. Birrento ML
   2. Bryan TM
   3. Samosorn S
   4. Beck JL

   (2015) ESI-MS investigation of an equilibrium between a bimolecular quadruplex DNA and a duplex DNA/RNA hybrid

   Journal of the American Society for Mass Spectrometry 26:1165–1173.

   https://doi.org/10.1007/s13361-015-1121-2

   • PubMed
   • Google Scholar
6. 1. Blackburn EH

   (1991) Structure and function of telomeres

   Nature 350:569–573.

   https://doi.org/10.1038/350569a0

   • PubMed
   • Google Scholar
7. 1. Bochman ML
   2. Paeschke K
   3. Zakian VA

   (2012) DNA secondary structures: stability and function of G-quadruplex structures

   Nature Reviews Genetics 13:770–780.

   https://doi.org/10.1038/nrg3296

   • PubMed
   • Google Scholar
8. Conference

   1. Bowers KJ
   2. Chow E
   3. Xu H
   4. Dror RO
   5. Eastwood MP
   6. Gregersen BA
   7. Klepeis JL
   8. Kolossvary I
   9. Moraes FD
   10. Sacerdoti MA
   11. Salmon JK
   12. Shan DE
   13. Shaw Y

   (2006)

   Scalable algorithms for molecular dynamics simulations on commodity clusters

   Proceedings of the 2006 ACM/IEEE Conference on Supercomputing.

   • Google Scholar
9. Software

   1. Bronson JE
   2. Fei J
   3. Hofman JM
   4. Gonzalez RL
   5. Wiggin CH

   (2013) vbFRET

   SourceForge.

   https://sourceforge.net/projects/vbfret/
10. 1. Budhathoki JB
    2. Maleki P
    3. Roy WA
    4. Janscak P
    5. Yodh JG
    6. Balci H

    (2016) A comparative study of G-Quadruplex unfolding and DNA reeling activities of human RECQ5 helicase

    Biophysical Journal 110:2585–2596.

    https://doi.org/10.1016/j.bpj.2016.05.016

    • PubMed
    • Google Scholar
11. 1. Burge S
    2. Parkinson GN
    3. Hazel P
    4. Todd AK
    5. Neidle S

    (2006) Quadruplex DNA: sequence, topology and structure

    Nucleic Acids Research 34:5402–5415.

    https://doi.org/10.1093/nar/gkl655

    • PubMed
    • Google Scholar
12. 1. Byrd AK
    2. Raney KD

    (2015) A parallel quadruplex DNA is bound tightly but unfolded slowly by pif1 helicase

    Journal of Biological Chemistry 290:6482–6494.

    https://doi.org/10.1074/jbc.M114.630749

    • PubMed
    • Google Scholar
13. 1. Chaires JB
    2. Gray RD
    3. Dean WL
    4. Monsen R
    5. DeLeeuw LW
    6. Stribinskis V
    7. Trent JO

    (2020) Human POT1 unfolds G-quadruplexes by conformational selection

    Nucleic Acids Research 48:4976–4991.

    https://doi.org/10.1093/nar/gkaa202

    • PubMed
    • Google Scholar
14. 1. Chen MC
    2. Tippana R
    3. Demeshkina NA
    4. Murat P
    5. Balasubramanian S
    6. Myong S
    7. Ferré-D'Amaré AR

    (2018) Structural basis of G-quadruplex unfolding by the DEAH/RHA helicase DHX36

    Nature 558:465–469.

    https://doi.org/10.1038/s41586-018-0209-9

    • PubMed
    • Google Scholar
15. 1. Chung WJ
    2. Heddi B
    3. Hamon F
    4. Teulade-Fichou MP
    5. Phan AT

    (2014) Solution structure of a G-quadruplex bound to the bisquinolinium compound Phen-DC(3)

    Angewandte Chemie International Edition 53:999–1002.

    https://doi.org/10.1002/anie.201308063

    • PubMed
    • Google Scholar
16. 1. Cohen SB
    2. Reddel RR

    (2008) A sensitive direct human telomerase activity assay

    Nature Methods 5:355–360.

    https://doi.org/10.1038/nmeth.f.209

    • PubMed
    • Google Scholar
17. 1. Counter CM
    2. Meyerson M
    3. Eaton EN
    4. Weinberg RA

    (1997) The catalytic subunit of yeast telomerase

    PNAS 94:9202–9207.

    https://doi.org/10.1073/pnas.94.17.9202

    • PubMed
    • Google Scholar
18. 1. Dai J
    2. Carver M
    3. Punchihewa C
    4. Jones RA
    5. Yang D

    (2007) Structure of the Hybrid-2 type intramolecular human telomeric G-quadruplex in K+ solution: insights into structure polymorphism of the human telomeric sequence

    Nucleic Acids Research 35:4927–4940.

    https://doi.org/10.1093/nar/gkm522

    • PubMed
    • Google Scholar
19. 1. De Cian A
    2. Cristofari G
    3. Reichenbach P
    4. De Lemos E
    5. Monchaud D
    6. Teulade-Fichou MP
    7. Shin-Ya K
    8. Lacroix L
    9. Lingner J
    10. Mergny JL

    (2007a) Reevaluation of telomerase inhibition by quadruplex ligands and their mechanisms of action

    PNAS 104:17347–17352.

    https://doi.org/10.1073/pnas.0707365104

    • PubMed
    • Google Scholar
20. 1. De Cian A
    2. Delemos E
    3. Mergny JL
    4. Teulade-Fichou MP
    5. Monchaud D

    (2007b) Highly efficient G-quadruplex recognition by bisquinolinium compounds

    Journal of the American Chemical Society 129:1856–1857.

    https://doi.org/10.1021/ja067352b

    • PubMed
    • Google Scholar
21. 1. Floyd DL
    2. Harrison SC
    3. van Oijen AM

    (2010) Analysis of kinetic intermediates in single-particle dwell-time distributions

    Biophysical Journal 99:360–366.

    https://doi.org/10.1016/j.bpj.2010.04.049

    • PubMed
    • Google Scholar
22. 1. Gavathiotis E
    2. Searle MS

    (2003) Structure of the parallel-stranded DNA quadruplex d(TTAGGGT)4 containing the human telomeric repeat: evidence for A-tetrad formation from NMR and molecular dynamics simulations

    Organic & Biomolecular Chemistry 1:1650–1656.

    https://doi.org/10.1039/b300845m

    • PubMed
    • Google Scholar
23. 1. Ghosh S
    2. Pradhan SK
    3. Kar A
    4. Chowdhury S
    5. Dasgupta D

    (2013) Molecular basis of recognition of quadruplexes human telomere and c-myc promoter by the putative anticancer agent sanguinarine

    Biochimica Et Biophysica Acta (BBA) - General Subjects 1830:4189–4201.

    https://doi.org/10.1016/j.bbagen.2013.03.027

    • PubMed
    • Google Scholar
24. 1. Gillis AJ
    2. Schuller AP
    3. Skordalakes E

    (2008) Structure of the tribolium castaneum telomerase catalytic subunit TERT

    Nature 455:633–637.

    https://doi.org/10.1038/nature07283

    • PubMed
    • Google Scholar
25. 1. Gomez DL
    2. Armando RG
    3. Cerrudo CS
    4. Ghiringhelli PD
    5. Gomez DE

    (2016) Telomerase as a Cancer target development of new molecules

    Current Topics in Medicinal Chemistry 16:2432–2440.

    https://doi.org/10.2174/1568026616666160212122425

    • PubMed
    • Google Scholar
26. 1. Gray DM
    2. Wen JD
    3. Gray CW
    4. Repges R
    5. Repges C
    6. Raabe G
    7. Fleischhauer J

    (2008) Measured and calculated CD spectra of G-quartets stacked with the same or opposite polarities

    Chirality 20:431–440.

    https://doi.org/10.1002/chir.20455

    • PubMed
    • Google Scholar
27. 1. Green JJ
    2. Ying L
    3. Klenerman D
    4. Balasubramanian S

    (2003) Kinetics of unfolding the human telomeric DNA quadruplex using a PNA trap

    Journal of the American Chemical Society 125:3763–3767.

    https://doi.org/10.1021/ja029149w

    • PubMed
    • Google Scholar
28. 1. Greider CW
    2. Blackburn EH

    (1985) Identification of a specific telomere terminal transferase activity in Tetrahymena extracts

    Cell 43:405–413.

    https://doi.org/10.1016/0092-8674(85)90170-9

    • PubMed
    • Google Scholar
29. 1. Greider CW
    2. Blackburn EH

    (1989) A telomeric sequence in the RNA of Tetrahymena telomerase required for telomere repeat synthesis

    Nature 337:331–337.

    https://doi.org/10.1038/337331a0

    • PubMed
    • Google Scholar
30. 1. Gueddouda NM
    2. Mendoza O
    3. Gomez D
    4. Bourdoncle A
    5. Mergny JL

    (2017) G-quadruplexes unfolding by RHAU helicase

    Biochimica Et Biophysica Acta (BBA) - General Subjects 1861:1382–1388.

    https://doi.org/10.1016/j.bbagen.2017.01.006

    • PubMed
    • Google Scholar
31. 1. Harley CB
    2. Futcher AB
    3. Greider CW

    (1990) Telomeres shorten during ageing of human fibroblasts

    Nature 345:458–460.

    https://doi.org/10.1038/345458a0

    • PubMed
    • Google Scholar
32. 1. Harrington L
    2. Zhou W
    3. McPhail T
    4. Oulton R
    5. Yeung DS
    6. Mar V
    7. Bass MB
    8. Robinson MO

    (1997) Human telomerase contains evolutionarily conserved catalytic and structural subunits

    Genes & Development 11:3109–3115.

    https://doi.org/10.1101/gad.11.23.3109

    • PubMed
    • Google Scholar
33. 1. Heddi B
    2. Phan AT

    (2011) Structure of human telomeric DNA in crowded solution

    Journal of the American Chemical Society 133:9824–9833.

    https://doi.org/10.1021/ja200786q

    • PubMed
    • Google Scholar
34. 1. Hiyama E
    2. Hiyama K

    (2007) Telomere and telomerase in stem cells

    British Journal of Cancer 96:1020–1024.

    https://doi.org/10.1038/sj.bjc.6603671

    • PubMed
    • Google Scholar
35. 1. Huppert JL
    2. Balasubramanian S

    (2005) Prevalence of quadruplexes in the human genome

    Nucleic Acids Research 33:2908–2916.

    https://doi.org/10.1093/nar/gki609

    • PubMed
    • Google Scholar
36. 1. Hwang H
    2. Buncher N
    3. Opresko PL
    4. Myong S

    (2012) POT1-TPP1 regulates telomeric overhang structural dynamics

    Structure 20:1872–1880.

    https://doi.org/10.1016/j.str.2012.08.018

    • PubMed
    • Google Scholar
37. 1. Hwang H
    2. Kreig A
    3. Calvert J
    4. Lormand J
    5. Kwon Y
    6. Daley JM
    7. Sung P
    8. Opresko PL
    9. Myong S

    (2014) Telomeric overhang length determines structural dynamics and accessibility to telomerase and ALT-associated proteins

    Structure 22:842–853.

    https://doi.org/10.1016/j.str.2014.03.013

    • PubMed
    • Google Scholar
38. 1. Ishikawa-Ankerhold HC
    2. Ankerhold R
    3. Drummen GP

    (2012) Advanced fluorescence microscopy techniques--FRAP, FLIP, FLAP, FRET and FLIM

    Molecules 17:4047–4132.

    https://doi.org/10.3390/molecules17044047

    • PubMed
    • Google Scholar
39. 1. Jansson LI
    2. Hentschel J
    3. Parks JW
    4. Chang TR
    5. Lu C
    6. Baral R
    7. Bagshaw CR
    8. Stone MD

    (2019) Telomere DNA G-quadruplex folding within actively extending human telomerase

    PNAS 116:9350–9359.

    https://doi.org/10.1073/pnas.1814777116

    • PubMed
    • Google Scholar
40. 1. Jiang J
    2. Wang Y
    3. Sušac L
    4. Chan H
    5. Basu R
    6. Zhou ZH
    7. Feigon J

    (2018) Structure of telomerase with telomeric DNA

    Cell 173:1179–1190.

    https://doi.org/10.1016/j.cell.2018.04.038

    • Google Scholar
41. 1. Kim NW
    2. Piatyszek MA
    3. Prowse KR
    4. Harley CB
    5. West MD
    6. Ho PL
    7. Coviello GM
    8. Wright WE
    9. Weinrich SL
    10. Shay JW

    (1994) Specific association of human telomerase activity with immortal cells and Cancer

    Science 266:2011–2015.

    https://doi.org/10.1126/science.7605428

    • PubMed
    • Google Scholar
42. 1. Lacroix L
    2. Séosse A
    3. Mergny J-L

    (2011) Fluorescence-based duplex–quadruplex competition test to screen for telomerase RNA quadruplex ligands

    Nucleic Acids Research 39:e21.

    https://doi.org/10.1093/nar/gkq1199

    • Google Scholar
43. 1. Lee HT
    2. Bose A
    3. Lee CY
    4. Opresko PL
    5. Myong S

    (2017) Molecular mechanisms by which oxidative DNA damage promotes telomerase activity

    Nucleic Acids Research 45:11752–11765.

    https://doi.org/10.1093/nar/gkx789

    • PubMed
    • Google Scholar
44. 1. Lim KW
    2. Amrane S
    3. Bouaziz S
    4. Xu W
    5. Mu Y
    6. Patel DJ
    7. Luu KN
    8. Phan AT

    (2009) Structure of the human telomere in K+ solution: a stable basket-type G-quadruplex with only two G-tetrad layers

    Journal of the American Chemical Society 131:4301–4309.

    https://doi.org/10.1021/ja807503g

    • PubMed
    • Google Scholar
45. 1. Lingner J
    2. Hughes TR
    3. Shevchenko A
    4. Mann M
    5. Lundblad V
    6. Cech TR

    (1997) Reverse transcriptase motifs in the catalytic subunit of telomerase

    Science 276:561–567.

    https://doi.org/10.1126/science.276.5312.561

    • PubMed
    • Google Scholar
46. 1. Liu JQ
    2. Chen CY
    3. Xue Y
    4. Hao YH
    5. Tan Z

    (2010) G-quadruplex hinders translocation of BLM helicase on DNA: a real-time fluorescence spectroscopic unwinding study and comparison with duplex substrates

    Journal of the American Chemical Society 132:10521–10527.

    https://doi.org/10.1021/ja1038165

    • PubMed
    • Google Scholar
47. 1. Liu HY
    2. Zhao Q
    3. Zhang TP
    4. Wu Y
    5. Xiong YX
    6. Wang SK
    7. Ge YL
    8. He JH
    9. Lv P
    10. Ou TM
    11. Tan JH
    12. Li D
    13. Gu LQ
    14. Ren J
    15. Zhao Y
    16. Huang ZS

    (2016) Conformation selective antibody enables genome profiling and leads to discovery of parallel G-Quadruplex in human telomeres

    Cell Chemical Biology 23:1261–1270.

    https://doi.org/10.1016/j.chembiol.2016.08.013

    • PubMed
    • Google Scholar
48. 1. London TB
    2. Barber LJ
    3. Mosedale G
    4. Kelly GP
    5. Balasubramanian S
    6. Hickson ID
    7. Boulton SJ
    8. Hiom K

    (2008) FANCJ is a structure-specific DNA helicase associated with the maintenance of genomic G/C tracts

    Journal of Biological Chemistry 283:36132–36139.

    https://doi.org/10.1074/jbc.M808152200

    • PubMed
    • Google Scholar
49. 1. Long X
    2. Stone MD

    (2013) Kinetic partitioning modulates human telomere DNA G-quadruplex structural polymorphism

    PLOS ONE 8:e83420.

    https://doi.org/10.1371/journal.pone.0083420

    • PubMed
    • Google Scholar
50. 1. Maizels N
    2. Gray LT

    (2013) The G4 genome

    PLOS Genetics 9:e1003468.

    https://doi.org/10.1371/journal.pgen.1003468

    • PubMed
    • Google Scholar
51. 1. Miller MC
    2. Buscaglia R
    3. Chaires JB
    4. Lane AN
    5. Trent JO

    (2010) Hydration is a major determinant of the G-quadruplex stability and conformation of the human telomere 3' sequence of d(AG3(TTAG3)3)

    Journal of the American Chemical Society 132:17105–17107.

    https://doi.org/10.1021/ja105259m

    • PubMed
    • Google Scholar
52. 1. Morin GB

    (1989) The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats

    Cell 59:521–529.

    https://doi.org/10.1016/0092-8674(89)90035-4

    • PubMed
    • Google Scholar
53. 1. Moye AL
    2. Porter KC
    3. Cohen SB
    4. Phan T
    5. Zyner KG
    6. Sasaki N
    7. Lovrecz GO
    8. Beck JL
    9. Bryan TM

    (2015) Telomeric G-quadruplexes are a substrate and site of localization for human telomerase

    Nature Communications 6:7643.

    https://doi.org/10.1038/ncomms8643

    • PubMed
    • Google Scholar
54. 1. Murat P
    2. Balasubramanian S

    (2014) Existence and consequences of G-quadruplex structures in DNA

    Current Opinion in Genetics & Development 25:22–29.

    https://doi.org/10.1016/j.gde.2013.10.012

    • PubMed
    • Google Scholar
55. 1. Nguyen THD
    2. Tam J
    3. Wu RA
    4. Greber BJ
    5. Toso D
    6. Nogales E
    7. Collins K

    (2018) Cryo-EM structure of substrate-bound human telomerase holoenzyme

    Nature 557:190–195.

    https://doi.org/10.1038/s41586-018-0062-x

    • PubMed
    • Google Scholar
56. 1. Nicoludis JM
    2. Barrett SP
    3. Mergny JL
    4. Yatsunyk LA

    (2012a) Interaction of human telomeric DNA with N-methyl mesoporphyrin IX

    Nucleic Acids Research 40:5432–5447.

    https://doi.org/10.1093/nar/gks152

    • PubMed
    • Google Scholar
57. 1. Nicoludis JM
    2. Miller ST
    3. Jeffrey PD
    4. Barrett SP
    5. Rablen PR
    6. Lawton TJ
    7. Yatsunyk LA

    (2012b) Optimized end-stacking provides specificity of N-methyl mesoporphyrin IX for human telomeric G-quadruplex DNA

    Journal of the American Chemical Society 134:20446–20456.

    https://doi.org/10.1021/ja3088746

    • PubMed
    • Google Scholar
58. 1. Oganesian L
    2. Moon IK
    3. Bryan TM
    4. Jarstfer MB

    (2006) Extension of G-quadruplex DNA by ciliate telomerase

    The EMBO Journal 25:1148–1159.

    https://doi.org/10.1038/sj.emboj.7601006

    • PubMed
    • Google Scholar
59. 1. Oganesian L
    2. Graham ME
    3. Robinson PJ
    4. Bryan TM

    (2007) Telomerase recognizes G-quadruplex and linear DNA as distinct substrates

    Biochemistry 46:11279–11290.

    https://doi.org/10.1021/bi700993q

    • PubMed
    • Google Scholar
60. 1. Olovnikov AM

    (1971)

    Principle of marginotomy in template synthesis of polynucleotides]

    Doklady Akademii Nauk SSSR 201:1496–1499.

    • PubMed
    • Google Scholar
61. 1. Paeschke K
    2. Simonsson T
    3. Postberg J
    4. Rhodes D
    5. Lipps HJ

    (2005) Telomere end-binding proteins control the formation of G-quadruplex DNA structures in vivo

    Nature Structural & Molecular Biology 12:847–854.

    https://doi.org/10.1038/nsmb982

    • PubMed
    • Google Scholar
62. 1. Paeschke K
    2. Juranek S
    3. Simonsson T
    4. Hempel A
    5. Rhodes D
    6. Lipps HJ

    (2008) Telomerase recruitment by the telomere end binding protein-beta facilitates G-quadruplex DNA unfolding in ciliates

    Nature Structural & Molecular Biology 15:598–604.

    https://doi.org/10.1038/nsmb.1422

    • PubMed
    • Google Scholar
63. 1. Palacký J
    2. Vorlíčková M
    3. Kejnovská I
    4. Mojzeš P

    (2013) Polymorphism of human telomeric quadruplex structure controlled by DNA concentration: a raman study

    Nucleic Acids Research 41:1005–1016.

    https://doi.org/10.1093/nar/gks1135

    • PubMed
    • Google Scholar
64. 1. Parks JW
    2. Stone MD

    (2014) Coordinated DNA dynamics during the human telomerase catalytic cycle

    Nature Communications 5:4146.

    https://doi.org/10.1038/ncomms5146

    • PubMed
    • Google Scholar
65. 1. Patrick EM
    2. Slivka JD
    3. Payne B
    4. Comstock MJ
    5. Schmidt JC

    (2020) Observation of processive telomerase catalysis using high-resolution optical tweezers

    Nature Chemical Biology 16:801–809.

    https://doi.org/10.1038/s41589-020-0478-0

    • Google Scholar
66. 1. Peng CG
    2. Damha MJ

    (2007) G-quadruplex induced stabilization by 2′-deoxy-2′-fluoro-d-arabinonucleic acids (2′F-ANA)

    Nucleic Acids Research 35:4977–4988.

    https://doi.org/10.1093/nar/gkm520

    • Google Scholar
67. 1. Petraccone L
    2. Malafronte A
    3. Amato J
    4. Giancola C

    (2012) G-quadruplexes from human telomeric DNA: how many conformations in PEG containing solutions?

    The Journal of Physical Chemistry B 116:2294–2305.

    https://doi.org/10.1021/jp209170v

    • PubMed
    • Google Scholar
68. 1. Podlevsky JD
    2. Chen JJ

    (2012) It all comes together at the ends: telomerase structure, function, and biogenesis

    Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 730:3–11.

    https://doi.org/10.1016/j.mrfmmm.2011.11.002

    • PubMed
    • Google Scholar
69. 1. Postberg J
    2. Tsytlonok M
    3. Sparvoli D
    4. Rhodes D
    5. Lipps HJ

    (2012) A telomerase-associated RecQ protein-like helicase resolves telomeric G-quadruplex structures during replication

    Gene 497:147–154.

    https://doi.org/10.1016/j.gene.2012.01.068

    • PubMed
    • Google Scholar
70. 1. Ray S
    2. Qureshi MH
    3. Malcolm DW
    4. Budhathoki JB
    5. Çelik U
    6. Balci H

    (2013) RPA-Mediated unfolding of systematically varying G-Quadruplex structures

    Biophysical Journal 104:2235–2245.

    https://doi.org/10.1016/j.bpj.2013.04.004

    • Google Scholar
71. 1. Renciuk D
    2. Kejnovská I
    3. Skoláková P
    4. Bednárová K
    5. Motlová J
    6. Vorlícková M

    (2009) Arrangements of human telomere DNA quadruplex in physiologically relevant K+ solutions

    Nucleic Acids Research 37:6625–6634.

    https://doi.org/10.1093/nar/gkp701

    • PubMed
    • Google Scholar
72. 1. Rhodes D
    2. Lipps HJ

    (2015) G-quadruplexes and their regulatory roles in biology

    Nucleic Acids Research 43:8627–8637.

    https://doi.org/10.1093/nar/gkv862

    • PubMed
    • Google Scholar
73. 1. Samosorn S
    2. Tanwirat B
    3. Muhamad N
    4. Casadei G
    5. Tomkiewicz D
    6. Lewis K
    7. Suksamrarn A
    8. Prammananan T
    9. Gornall KC
    10. Beck JL
    11. Bremner JB

    (2009) Antibacterial activity of berberine-NorA pump inhibitor hybrids with a methylene ether linking group

    Bioorganic & Medicinal Chemistry 17:3866–3872.

    https://doi.org/10.1016/j.bmc.2009.04.028

    • PubMed
    • Google Scholar
74. Patent

    1. Samosorn S

    (2016) 13-Arylalkyleneoxyberberine deriavatives as Anti-Breast Cancer agents (TP 11579)

    Patent. Thailand.

    https://patentsearch.ipthailand.go.th/DIP2013/view_public_data.php?appno=11125900022
75. 1. Schaffitzel C
    2. Berger I
    3. Postberg J
    4. Hanes J
    5. Lipps HJ
    6. Plückthun A

    (2001) In vitro generated antibodies specific for telomeric guanine-quadruplex DNA react with Stylonychia lemnae macronuclei

    PNAS 98:8572–8577.

    https://doi.org/10.1073/pnas.141229498

    • PubMed
    • Google Scholar
76. 1. Sen D
    2. Gilbert W

    (1988) Formation of parallel four-stranded complexes by guanine-rich motifs in DNA and its implications for meiosis

    Nature 334:364–366.

    https://doi.org/10.1038/334364a0

    • PubMed
    • Google Scholar
77. 1. Sindbert S
    2. Kalinin S
    3. Nguyen H
    4. Kienzler A
    5. Clima L
    6. Bannwarth W
    7. Appel B
    8. Müller S
    9. Seidel CA

    (2011) Accurate distance determination of nucleic acids via förster resonance energy transfer: implications of dye Linker length and rigidity

    Journal of the American Chemical Society 133:2463–2480.

    https://doi.org/10.1021/ja105725e

    • PubMed
    • Google Scholar
78. 1. Smith JS
    2. Chen Q
    3. Yatsunyk LA
    4. Nicoludis JM
    5. Garcia MS
    6. Kranaster R
    7. Balasubramanian S
    8. Monchaud D
    9. Teulade-Fichou MP
    10. Abramowitz L
    11. Schultz DC
    12. Johnson FB

    (2011) Rudimentary G-quadruplex-based telomere capping in Saccharomyces cerevisiae

    Nature Structural & Molecular Biology 18:478–485.

    https://doi.org/10.1038/nsmb.2033

    • PubMed
    • Google Scholar
79. 1. Sun D
    2. Thompson B
    3. Cathers BE
    4. Salazar M
    5. Kerwin SM
    6. Trent JO
    7. Jenkins TC
    8. Neidle S
    9. Hurley LH

    (1997) Inhibition of human telomerase by a G-quadruplex-interactive compound

    Journal of Medicinal Chemistry 40:2113–2116.

    https://doi.org/10.1021/jm970199z

    • PubMed
    • Google Scholar
80. 1. Sundquist WI
    2. Klug A

    (1989) Telomeric DNA dimerizes by formation of guanine tetrads between hairpin loops

    Nature 342:825–829.

    https://doi.org/10.1038/342825a0

    • PubMed
    • Google Scholar
81. 1. Syed S
    2. Pandey M
    3. Patel SS
    4. Ha T

    (2014) Single-molecule fluorescence reveals the unwinding stepping mechanism of replicative helicase

    Cell Reports 6:1037–1045.

    https://doi.org/10.1016/j.celrep.2014.02.022

    • PubMed
    • Google Scholar
82. 1. Tippana R
    2. Hwang H
    3. Opresko PL
    4. Bohr VA
    5. Myong S

    (2016) Single-molecule imaging reveals a common mechanism shared by G-quadruplex-resolving helicases

    PNAS 113:8448–8453.

    https://doi.org/10.1073/pnas.1603724113

    • PubMed
    • Google Scholar
83. 1. Tomlinson CG
    2. Moye AL
    3. Holien JK
    4. Parker MW
    5. Cohen SB
    6. Bryan TM

    (2015) Two-step mechanism involving active-site conformational changes regulates human telomerase DNA binding

    Biochemical Journal 465:347–357.

    https://doi.org/10.1042/BJ20140922

    • PubMed
    • Google Scholar
84. 1. Tomlinson CG
    2. Sasaki N
    3. Jurczyluk J
    4. Bryan TM
    5. Cohen SB

    (2017) Quantitative assays for measuring human telomerase activity and DNA binding properties

    Methods 114:85–95.

    https://doi.org/10.1016/j.ymeth.2016.08.002

    • PubMed
    • Google Scholar
85. 1. Wang F
    2. Tang ML
    3. Zeng ZX
    4. Wu RY
    5. Xue Y
    6. Hao YH
    7. Pang DW
    8. Zhao Y
    9. Tan Z

    (2012) Telomere- and telomerase-interacting protein that unfolds telomere G-quadruplex and promotes telomere extension in mammalian cells

    PNAS 109:20413–20418.

    https://doi.org/10.1073/pnas.1200232109

    • PubMed
    • Google Scholar
86. 1. Waterhouse A
    2. Bertoni M
    3. Bienert S
    4. Studer G
    5. Tauriello G
    6. Gumienny R
    7. Heer FT
    8. de Beer TAP
    9. Rempfer C
    10. Bordoli L
    11. Lepore R
    12. Schwede T

    (2018) SWISS-MODEL: homology modelling of protein structures and complexes

    Nucleic Acids Research 46:W296–W303.

    https://doi.org/10.1093/nar/gky427

    • PubMed
    • Google Scholar
87. 1. Watson JD

    (1972) Origin of concatemeric T7 DNA

    Nature New Biology 239:197–201.

    https://doi.org/10.1038/newbio239197a0

    • PubMed
    • Google Scholar
88. 1. Weinrich SL
    2. Pruzan R
    3. Ma L
    4. Ouellette M
    5. Tesmer VM
    6. Holt SE
    7. Bodnar AG
    8. Lichtsteiner S
    9. Kim NW
    10. Trager JB
    11. Taylor RD
    12. Carlos R
    13. Andrews WH
    14. Wright WE
    15. Shay JW
    16. Harley CB
    17. Morin GB

    (1997) Reconstitution of human telomerase with the template RNA component hTR and the catalytic protein subunit hTRT

    Nature Genetics 17:498–502.

    https://doi.org/10.1038/ng1297-498

    • PubMed
    • Google Scholar
89. 1. Williamson JR
    2. Raghuraman MK
    3. Cech TR

    (1989) Monovalent cation-induced structure of telomeric DNA: the G-quartet model

    Cell 59:871–880.

    https://doi.org/10.1016/0092-8674(89)90610-7

    • PubMed
    • Google Scholar
90. 1. Wu RA
    2. Tam J
    3. Collins K

    (2017) DNA-binding determinants and cellular thresholds for human telomerase repeat addition processivity

    The EMBO Journal 36:1908–1927.

    https://doi.org/10.15252/embj.201796887

    • PubMed
    • Google Scholar
91. 1. Wyatt HD
    2. Lobb DA
    3. Beattie TL

    (2007) Characterization of physical and functional anchor site interactions in human telomerase

    Molecular and Cellular Biology 27:3226–3240.

    https://doi.org/10.1128/MCB.02368-06

    • PubMed
    • Google Scholar
92. 1. Wyatt HD
    2. West SC
    3. Beattie TL

    (2010) InTERTpreting telomerase structure and function

    Nucleic Acids Research 38:5609–5622.

    https://doi.org/10.1093/nar/gkq370

    • PubMed
    • Google Scholar
93. 1. Yang W
    2. Lee YS

    (2015) A DNA-hairpin model for repeat-addition processivity in telomere synthesis

    Nature Structural & Molecular Biology 22:844–847.

    https://doi.org/10.1038/nsmb.3098

    • PubMed
    • Google Scholar
94. 1. Zahler AM
    2. Williamson JR
    3. Cech TR
    4. Prescott DM

    (1991) Inhibition of telomerase by G-quartet DNA structures

    Nature 350:718–720.

    https://doi.org/10.1038/350718a0

    • PubMed
    • Google Scholar
95. 1. Zaug AJ
    2. Podell ER
    3. Cech TR

    (2005) Human POT1 disrupts telomeric G-quadruplexes allowing telomerase extension in vitro

    PNAS 102:10864–10869.

    https://doi.org/10.1073/pnas.0504744102

    • PubMed
    • Google Scholar
96. 1. Zhang ML
    2. Tong XJ
    3. Fu XH
    4. Zhou BO
    5. Wang J
    6. Liao XH
    7. Li QJ
    8. Shen N
    9. Ding J
    10. Zhou JQ

    (2010) Yeast telomerase subunit Est1p has guanine quadruplex-promoting activity that is required for telomere elongation

    Nature Structural & Molecular Biology 17:202–209.

    https://doi.org/10.1038/nsmb.1760

    • PubMed
    • Google Scholar

Thank you for submitting your article "A mechanism for the extension and unfolding of parallel G-quadruplexes by human telomerase at single-molecule resolution" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by John Kuriyan as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper.

Summary:

Telomerase is the enzyme tasked with maintaining telomeres at the ends of linear eukaryotic chromosomes. Telomere DNA sequence is repetitive and guanine rich in most organisms, and the tandem arrays of guanine bases can form multiple planar G-quartet motifs that are stabilized by Hoogsteen hydrogen bonding. In turn, several G-quartets can stack upon each other to form G-quadruplex (GQ) DNA folds, characterized by their very polymorphic structural properties and high melting temperatures. This manuscript by the van Oijen and Bryan groups builds on previous work by the Bryan lab where they showed that purified human telomerase disrupts and processively extends parallel intermolecular G4 structures. In this current manuscript, the authors present a new smFRET-based approach to observe the dynamics and intermediates during telomerase-mediated resolution of parallel G4 structures. Combined with bulk biochemistry, these measurements establish a three-step mechanism for the unfolding and resolution of parallel intermolecular G4 structures.

This manuscript provides significant new mechanistic insight into the function of human telomerase, and the single-molecule data and analyses are overall of excellent quality. A subset of the conclusions seems to be overstated; here the manuscript would benefit from a toned-down interpretation of the data presented (see revisions). Please note that in response to the current COVID-19 situation, only textual changes are required, however already existing data addressing the points raised below should be included, if available.

Essential revisions:

1) The FRET values in the study are interpreted quite loosely. For example, there is no control experiment presented that demonstrates the low FRET state observed in their experiments is in fact a completely resolved GQ. While the interpretation for this experiment is certainly reasonable based upon previous smFRET studies of GQ structures, the authors should ideally substantiate this assignment with mutant DNA substrates or perhaps using Li+ in their buffers. If such new data cannot currently be acquired, the authors should better substantiate their claim that the low FRET state is unfolded by more explicitly referring to other published work. The assignment of all FRET states in this paper would additionally benefit from some structural modeling to estimate expected distances.

2) Related to the first point, it is surprising that both the intra- and intermolecular GQ FRET experiments give essentially the same signal? The interpretation of the two-step drop seems plausible for the intramolecular GQ, but for the intermolecular structure it is unclear why initial docking on one of the four 3' ends would impact the positioning of the FRET dyes located on the 5' ends of the DNA substrate (as drawn in Figure 2A)? Also, is ~0.55 FRET a reasonable value for the schematic drawings shown in Figure 2A, in other words what is the known distance between dyes on neighboring strands of an intermolecular GQ? One would expect this to be quite close (~2-3 nm) and therefore give rise to a much higher initial FRET state. It is also unclear why telomerase binding and remodeling of this tetrameric structure would not cause the structure to fall apart, and therefore give rise to a zero FRET state or loss of signal altogether. Is it possible this is happening and just not being detected in these experiments? The authors should address this in the text.

3) Given the results from Hwang, Myong et al., 2012 (not cited) on the binding and resolution of GQ structure by POT1-TPP1 protein, which occurs in three steps, is it unexpected that two steps is what is observed in the present work? The authors should discuss their results in light of this earlier work.

4) The authors' use of active versus passive unwinding mechanisms for telomerase-mediated GQ unfolding is somewhat unclear. The argument for 'active' unwinding is based upon experiments performed with a short RNA oligo presented in Figure 3 and Supplementary Figure 3, in which this short RNA is meant to mimic the RNA template within the assembled telomerase RNP. The authors state that given the concentration dependence of the RNA-induced FRET drop, the mechanism must therefore be 'active' – but why is this necessarily true? The authors do not show any kinetic data for this experiment, therefore it is difficult to ascertain whether the process displays zero or first-order kinetics with respect to oligo-induced disruption of the GQ structure. Moreover, since there is no energy cofactor (i.e., dNTP) in this experiment, is not the mechanism, by definition, passive? The authors should articulate specifically how they are using the word 'active' in this section of the study? In addition, the authors seem to think the experiments in this figure rule out the possibility of transient intrinsic GQ dynamics being 'captured' by the RNA oligo (or telomerase RNP). However, previous work including that by Green et al., 2003 (not cited) describes a similar experiment with PNA oligos and comes to the opposite conclusion. Also, it seems unlikely that a 10 nt RNA oligo fully captures the properties of the RNA template within the telomerase RNP context – so a more convincing experiment to support the conclusion that it is just the RNA template that is required for the first FRET drop would be to use either a mutant template RNA and/or mutant DNA substrate. At the very least, the authors should discuss the limitations of using a 10 nt RNA oligo to mimic the properties of the telomerase RNA template within the assembled complex.

5) In Figure 4, the authors present very interesting smFRET data for the system being 'walked' through the telomere DNA synthesis reaction through the use of ddNTP terminators. This experiment, and indeed aspects of the study more generally are similar to work from Parks and Stone, 2014 (not cited), in which DNA primer dynamics are directly observed by smFRET in human telomerase as the primer length approaches the template boundary. Ultimately, the authors claim that the experiment in Figure 4I argues that telomerase translocation is required to fully disrupt the GQ structure, evidenced by the second FRET drop upon incorporation of a ddGTP. The conclusion that translocation leads to GQ unfolding is too strong given the data, since in this experiment there is no evidence of translocation having occurred. Typically, translocation would be thought of as being complete once the necessary rearrangements required to allow re-priming for subsequent rounds of telomere DNA synthesis have occurred – but it is not evident this is happening in the experiments described since there is a chain terminator. Therefore, the authors should revise their use of the term translocation so as to avoid an over-interpretation of their very interesting data.

6) In Figures 5 and 6 the effects of GQ binding ligands are studied and the conclusion is that these compounds do not completely inhibit telomerase in the smFRET assay. Why are the results of the gel-based assay for the F-22G3 and the smFRET experiments not more similar? It appears in Figure 5L that NMM and PhenDC3 almost completely inhibit the telomerase activity, whereas SST16 does not have as strong of an effect (the authors should quantify the relative activities in this experiment and include them as part of the figure). In contrast, all three compounds yield similar results in the smFRET assay as shown in the bar plots in Figure 5J. The authors should explain this apparent discrepancy, as this might seem to imply that the two-step FRET drop is not correlating with catalytic activity in this experiment. The authors also comment on the impact of PhenDC3 causing a change in the product profile after four repeats 'most likely resulting from G-quadruplex stabilization within the product DNA'. This result is in close accord with the recent study by Jansson et al., 2019 (not cited) wherein the possibility of GQ folding within the product of human telomerase during catalysis is described. In addition, in the Discussion, the authors relate the thermal stabilization of the structure by these compounds as evidence for 'telomerase not simply exploiting inherent instability in the G4 structure…' However, smFRET experiments by Jena, Ha et al. JACS 2009 (not cited) – showed that compounds that increase the thermal stability of the GQ fold do not necessarily restrict the conformational dynamics of the structure.

7) While the two-step FRET signal observed throughout the study is very interesting, the model put forth in Figure 7 is not very satisfying. Throughout the paper (including in the title) the authors describe how these data lead them to a mechanism for how telomerase acts as a GQ resolvase. However, the authors do not make an effort to discuss their data within a structural context based upon the medium-resolution cryo-EM structure of human telomerase (Nguyen et al., 2018, not cited) or the higher-resolution cryo-EM structure of Tetrahymena telomerase (Jiang et al., 2018, not cited). Why is there no attempt to explain how the proposed GQ unfolding might occur? For example, specifically how steric hindrance by the enzyme would cause unfolding during catalysis, but still allow binding to occur at the start of catalysis? Moreover, in the case of the tetrameric GQ, why would activity cause the structure to disassemble as would be suggested by the cartoon in Figure 7 (see comment above)? The authors should clarify this in the text.

8) The dwell time distribution shown in Figure 1N does not seem to be described particularly well by a single exponential decay. Nonetheless the authors argue that the distribution indicates only one rate-limiting step. Given that the observed reaction is a multi-step process, a more complex kinetic model might be required. If the authors have data recorded at various different dNTP concentrations, they should be included and discussed.

Related to this point, the second FRET decrease observed with different G4 structures, appears to occur abruptly/instantaneously. Can intermediates during the transition be visualized (at higher frame rates or limiting dNTP concentrations)? If the transition times can be quantified, what is the underlying distribution?

9) Have EMSAs or similar experiments been performed to clarify that the RNA is actually binding / invading the G4, whereas nothing happens with DNA? Similarly, it would be informative to see binding data (Kd) for telomerase to the G4 structures used here. Also, is the initial FRET change from 0.5 to 0.3 correlated in time with telomerase binding? In other words, does the time lag from injection to FRET shift match the telomerase on-rate, or does the FRET change require more time (indicating a more complex process)?

10) Is the FRET change observed by telomerase / RNA binding reversible, e.g. if telomerase dissociates? Similarly, given that very high concentrations of RNA were required to observe the effect seen in Figure 3F, RNA is most likely only loosely bound. It would be interesting to see if the G4 reverts when the RNA dissociates.

11) The authors find that G4-stabilizing ligands double the number of G4 that are not opened. This is not really discussed. Could it be that the G4s that are actually unwound are not really bound by the ligands? Does this effect depend on the ligand concentration?

There is also an apparent discrepancy between the amount of G4s that are stabilized, the change in unfolding rate (Figure 5K), and the extension rate (Figure 5L). For example, PhenDC3 results in a minor slowing down of G4 unfolding in Figure 5K but seems to completely inhibit extension (Figure 5L). These discrepancies should be discussed.

Essential revisions:

1) The FRET values in the study are interpreted quite loosely. For example, there is no control experiment presented that demonstrates the low FRET state observed in their experiments is in fact a completely resolved GQ. While the interpretation for this experiment is certainly reasonable based upon previous smFRET studies of GQ structures, the authors should ideally substantiate this assignment with mutant DNA substrates or perhaps using Li+ in their buffers. If such new data cannot currently be acquired, the authors should better substantiate their claim that the low FRET state is unfolded by more explicitly referring to other published work. The assignment of all FRET states in this paper would additionally benefit from some structural modeling to estimate expected distances.

We agree that controls showing the FRET from a denatured or resolved G4 are important. Since we were able to carry out a small amount of experimental work despite coronavirus restrictions, we performed the suggested experiments. After placing each of the two G4 constructs on the surface in potassium-based buffer, we replaced the potassium buffer with Li⁺- buffer and monitored the FRET signal. The data show that both G4s exhibited a drop in FRET signal to a value similar to that observed in the presence of telomerase and dNTPs (see Figure 2—figure supplement 1 and Figure 4—figure supplement 2), supporting the proposition that this FRET value is indicative of an unfolded G4.

The intramolecular G4 (F-22G3) FRET decreased from ∼0.5 to ∼0.14 (new Figure 2—figure supplement 1; subsection “Human telomerase binds, unfolds and extends intramolecular parallel G-quadruplexes”) and a majority of molecules remained at that lowest FRET state during the entire observation window. A minor proportion (35 ± 8%) of molecules showed dynamic transitioning between FRET states but favored the lowest FRET state.

Intermolecular G4 (F-7GGT) also displayed a FRET decrease from ∼0.6 to ∼0.2 after swapping K⁺ for Li⁺ (new Figure 4—figure supplement 2; subsection “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”), followed by a loss of FRET altogether, suggesting the G4 structure dissociates from the surface. This behaviour is readily explained by the fact that the biotinylated strand does not bear any fluorophores and that loss of the other strands causes the FRET signal to disappear. The presence of a stable FRET signal at ~0.2 in a majority of molecules prior to disappearance of the FRET signal suggests that the 0.2 FRET state represents an unwound G4, prior to complete diffusion of the non-biotinylated strands. As predicted by one of the reviewers, this behaviour does differ from that in the presence of telomerase + dNTPs (discussed further below), suggesting that the presence of telomerase prevents full dispersion of the unwound strands, and we have addressed this in the aforementioned subsection.

Overall, these results are consistent with our original interpretation whereby G4 extension by telomerase in the presence of dNTPs causes a decrease in FRET due to disruption of the G4 structure.

2) Related to the first point, it is surprising that both the intra- and intermolecular GQ FRET experiments give essentially the same signal?

The similar FRET values between the constructs is probably nothing more than a coincidence. The intra-molecular G4 exhibits FRET centred around 0.5, while inter-molecular has a FRET value centred around 0.6. To better understand the relationship between the expected FRET and the dye positions within the G4, we have now used previously published crystal (intramolecular G4) and NMR structures (intermolecular G4) to estimate the distance between dyes, respectively.

Importantly, we should note that reliable distance estimates obtained from FRET are extremely challenging, with the FRET efficiency depending upon several factors: 1) The length of linkers that are used to attach each fluorescent dye (in our case a C6 linker) will affect dye mobility and increase the distance between dyes, both processes affecting FRET. 2) The quenching of emission due to the presence of Gs in the dye’s vicinity. 3) Solvent polarity affecting FRET transfer (see AO Crockett and CT Wittwer, Anal Biochem [2001] vol 290: 89; S Sindbert et al., 2011; Ishikawa-Ankerhold, Ankerhold and Drummen, 2012).

The above factors notwithstanding, we calculated the FRET efficiency based on an available crystal structure for a version of F-22G3 without the 2'F-araG substitutions (parallel intramolecular GQ, PDB 1kf1). The estimated distance between the two dyes, which are located on the two strands as shown in Author response image 1, is approximately 53 Å including the length of the linkers. Given a Förster radius between AF555 and AF647 of 51 Å, the calculated FRET efficiency for this distance is 0.44, which is close to our experimental FRET value of 0.5. This similarity suggests there is not a significant change in dye emission in this construct due to quenching, with the acceptor dye being located away from the quenching Gs (see Author response image 1).

Author response image 1

Download asset Open asset

For the intermolecular GQ (F-7GGT), however, a high FRET efficiency is expected as the dyes are attached at the proximal 5’ ends, as pointed out by the reviewers. The estimated distance between dyes is approximately 25 – 35 Å, based on an NMR solution structure of this tetrameric G4 (PDB 1PN9; there is no crystal structure available), together with C-6 carbon linkers for fluorophore attachment. This would yield a theoretical FRET efficiency of ≥ 0.9; however, most of the molecules exhibited FRET efficiency distributed around 0.5 to 0.7 in our experiments. The discrepancy in measured experimental FRET compared to expected FRET suggests that fluorophore quenching occurs in this construct as both members of the FRET pair are close to the G-tetrad plane. Nevertheless, we are confident that the observed ~0.6 FRET state represents a parallel tetrameric G4, since i) there is no other G4 conformation that can form from this short oligonucleotide, and ii) the Li⁺ control above demonstrates a reduction in this signal upon G4 unwinding.

To our knowledge, there are no smFRET studies carried out using intermolecular or intramolecular parallel telomeric G4 constructs to directly compare with our FRET design.

With the caveat that relating FRET values to distances is challenging, we have briefly discussed these points in the revised manuscript (subsections “Human telomerase binds, unfolds and extends intramolecular parallel G-quadruplexes” and “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”).

The interpretation of the two-step drop seems plausible for the intramolecular GQ, but for the intermolecular structure it is unclear why initial docking on one of the four 3' ends would impact the positioning of the FRET dyes located on the 5' ends of the DNA substrate (as drawn in Figure 2A)?

We postulate that upon telomerase binding to intermolecular G4, since the template RNA of telomerase needs to hybridise with the 3’ end of one of the four strands of the G4, this will trigger a change in conformation (maybe flipping of the strand) in such a way that two fluorophores move further apart, resulting in a decrease in FRET. We have now performed molecular dynamic simulations of this hybridisation and potential strand flipping, in the context of a homology model of human telomerase (discussed further in response to point 7 below), and we do indeed observe the 5’ ends of the G4 move further apart between 10 ns and 100 ns of the simulation (see Figure 10B).

Also, is ~0.55 FRET a reasonable value for the schematic drawings shown in Figure 2A, in other words what is the known distance between dyes on neighboring strands of an intermolecular GQ? One would expect this to be quite close (~2-3 nm) and therefore give rise to a much higher initial FRET state.

This point is discussed above; the difference between observed and expected FRET strongly suggests the fluorophores are quenched partly by the presence of G-tetrad plane and other experimental conditions.

It is also unclear why telomerase binding and remodeling of this tetrameric structure would not cause the structure to fall apart, and therefore give rise to a zero FRET state or loss of signal altogether.

Yes, we agree with this point. Surprisingly, in our telomerase extension experiments, all molecules remained at a low FRET state before they photobleached. In contrast, when experiments were conducted in the presence of Li⁺- buffer, a complete loss of FRET signal was observed, suggesting complete dissociation of the non-tethered strands from the surface of attachment. This suggests that the presence of telomerase holds the strands of the G4 near each other for a longer period of time, whereas in the absence of telomerase the unwound strands are free to diffuse away. This has been discussed in the revised manuscript (subsection “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”).

Is it possible this is happening and just not being detected in these experiments? The authors should address this in the text.

Yes, this is one possibility. However, in our experiments in K⁺-buffer in the presence of telomerase, we observed that the majority of molecules displayed acceptor dye photobleaching followed by photobleaching of the donor dye. If the strands completely dissociated, one would expect the disappearance of both signals at the same time, as occurred in the Li⁺ control. This point has been discussed in the revised manuscript (subsection “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”).

3) Given the results from Hwang, Myong et al., 2012 (not cited) on the binding and resolution of GQ structure by POT1-TPP1 protein, which occurs in three steps, is it unexpected that two steps is what is observed in the present work? The authors should discuss their results in light of this earlier work.

Hwang et al., 2012, showed that POT1 unfolds a 24 nt G4 in four steps, whereas the POT1-TPP1 complex caused a 2-step FRET decrease followed by dynamic folding and unfolding. The mechanism of GQ unfolding by POT1 and the POT1-TPP1 complex is very different to that of telomerase; our data suggest that telomerase can directly hybridize to G4 via its RNA template, whereas POT1 unfolds G4 by a conformational selection mechanism in which G4 unfolding occurs prior to POT1 binding (Chaires et al., 2020). The four different steps of FRET reduction in the case of POT1 are attributable to the two OB folds of each of two POT1 molecules binding to the four “arms” of a G-quadruplex sequentially as it unfolds (Hwang et al., 2012). In our case, our model is that the first step of FRET reduction arises due to a conformational change upon RNA hybridization to the 3’ end of the G4, and the second step is due to complete unwinding upon telomerase extension. We have added mention of the POT1 results in Hwang et al., 2012 to the paragraph in the Discussion comparing the unique mechanism of telomerase G4 unwinding with that of other proteins.

4) The authors' use of active versus passive unwinding mechanisms for telomerase-mediated GQ unfolding is somewhat unclear. The argument for 'active' unwinding is based upon experiments performed with a short RNA oligo presented in Figure 3 and Supplementary Figure 3, in which this short RNA is meant to mimic the RNA template within the assembled telomerase RNP. The authors state that given the concentration dependence of the RNA-induced FRET drop, the mechanism must therefore be 'active' – but why is this necessarily true? The authors do not show any kinetic data for this experiment, therefore it is difficult to ascertain whether the process displays zero or first-order kinetics with respect to oligo-induced disruption of the GQ structure.

We used the term “passive” to mean a mechanism in which the enzyme traps an unfolded G4 in an equilibrium population, i.e. binding of the enzyme to the DNA is preceded by spontaneous G-quadruplex unwinding (as in the case of POT1; Chaires et al., 2020). “Active” referred to any mechanism that is NOT passive trapping, involving direct binding of telomerase to the G-quadruplex. However, the reviewer is correct that we did not measure the kinetics of RNA oligo-induced G4 unwinding. Unfortunately, because of the nonlinear nature of flow in the microfluidic flow cells, it is not possible to accurately measure the time between sample injection and the FRET drop. First, the starting times are measured from the time the RNA is injected. This is followed by a significant delay time (probably most of the delay time we see) before it physically enters the flow cell. Because of fluid dynamics, the RNA will not arrive in a coherent manner at the slide surface, but rather will show up as a gradually increasing concentration, making it impossible to measure binding time in an unambiguous manner. Furthermore, the microfluidic device varies from day to day in ways that affect flow rate. Therefore, it is not practical to measure meaningful initial binding kinetics. We are therefore unable to determine whether the concentration dependence of RNA-induced unwinding is first-order with respect to the RNA, so we have removed the sentence “the concentration-dependent unfolding by the RNA oligonucleotide (Supplementary Figure 3B, C) also supports an active invasion mechanism”.

Nevertheless, we believe that the fact that no FRET drop is observed in the presence of the same concentration of a DNA oligo is still strong evidence that RNA-mediated unfolding is NOT due to passive trapping of an unfolded G4, since this would also occur with a DNA oligo.

Moreover, since there is no energy cofactor (i.e., dNTP) in this experiment, is not the mechanism, by definition, passive? The authors should articulate specifically how they are using the word 'active' in this section of the study?

As discussed above, we were using “passive” to indicate a mechanism involving passive trapping of a spontaneously-unwound G4, rather than a mechanism that does not need the energy of nucleotide hydrolysis. There are other examples of G4-unwinding proteins whose mechanism involves neither passive trapping nor ATP hydrolysis. For example, the ability of RPA to unwind G4 is independent of the stability of the G4, and does not need ATP; the mechanism is thought to involve RPA binding to ssDNA loops in the G4 structure (Qureshi et al. J Phys Chem 2012; Ray et al., 2013). Nevertheless, to avoid ambiguity we have more extensively defined what we mean by a “passive” mechanism in the text (Discussion), and we have removed the term “active unwinding”.

In addition, the authors seem to think the experiments in this figure rule out the possibility of transient intrinsic GQ dynamics being 'captured' by the RNA oligo (or telomerase RNP). However, previous work including that by Green et al., 2003 (not cited) describes a similar experiment with PNA oligos and comes to the opposite conclusion.

The study by Green et al., 2003, (and most other previous work) used a different G4 to those in our study. They measured the kinetics of PNA probe binding to an antiparallel telomeric G4 folded in 100 mM NaCl, a quadruplex known to have relatively low stability (T_m of 47°C in their study, and see Balagurumoorthy and Brahmachari, JBC, 1994). They found that the unfolding was zero order with respect to the PNA oligo, suggesting that the initial step is a rate-determining internal rearrangement of the quadruplex. In contrast, the parallel G4 we used in this study were highly stable and showed NO unwinding over the time course of our smFRET experiments. Furthermore, a DNA oligo did not demonstrate ANY ability to trap the G4 structure within the time of our experiments. Conversely, RNA binding to G4 was observed within a few minutes. If there were transient intrinsic G4 dynamics, a DNA oligo would also have been able to capture the unfolded form. This point is now clarified in the revised manuscript (Discussion).

Also, it seems unlikely that a 10 nt RNA oligo fully captures the properties of the RNA template within the telomerase RNP context – so a more convincing experiment to support the conclusion that it is just the RNA template that is required for the first FRET drop would be to use either a mutant template RNA and/or mutant DNA substrate. At the very least, the authors should discuss the limitations of using a 10 nt RNA oligo to mimic the properties of the telomerase RNA template within the assembled complex.

We fully agree with this point; the telomerase RNP complex would have far more points of contact with the G4 than the RNA oligo, which is likely the reason that a high concentration of oligo was needed to see unwinding. We have not had the opportunity to perform any experiments with mutant version of those oligos, but we have edited the text to make the point that a 10 nt RNA oligo does not fully mimic the properties of the telomerase RNA template within the assembled complex (subsection “The RNA template sequence is involved in partial unfolding of G-quadruplex structure” and Discussion).

5) In Figure 4, the authors present very interesting smFRET data for the system being 'walked' through the telomere DNA synthesis reaction through the use of ddNTP terminators. This experiment, and indeed aspects of the study more generally are similar to work from Parks and Stone, 2014 (not cited), in which DNA primer dynamics are directly observed by smFRET in human telomerase as the primer length approaches the template boundary. Ultimately, the authors claim that the experiment in Figure 4I argues that telomerase translocation is required to fully disrupt the GQ structure, evidenced by the second FRET drop upon incorporation of a ddGTP. The conclusion that translocation leads to GQ unfolding is too strong given the data, since in this experiment there is no evidence of translocation having occurred. Typically, translocation would be thought of as being complete once the necessary rearrangements required to allow re-priming for subsequent rounds of telomere DNA synthesis have occurred – but it is not evident this is happening in the experiments described since there is a chain terminator. Therefore, the authors should revise their use of the term translocation so as to avoid an over-interpretation of their very interesting data.

The data in Parks and Stone, 2014, demonstrate that addition of a ddGTP opposite the end of the telomerase template sequence is followed by realignment of the DNA primer and rehybridization with the upstream region of the template (Figure 5B in their paper). Therefore, chain termination does not necessarily preclude translocation and rehybridization. There is no reason to believe that this part of the translocation process is not happening in our system also, since the G-quadruplex substrate is compatible with processive addition (new Figures 1 and 3 of our manuscript), and we have now referenced Parks and Stone, 2014, as evidence that it is likely the process of DNA realignment during translocation that triggers the FRET change we see (Discussion). However, the reviewer is correct that we don’t have direct evidence for this in our own experiments, so we have revised the language around our conclusions, now stating that the data in Figure 4 (now Figure 6) demonstrate that the FRET drop is triggered by synthesis reaching the end of the template, and is likely to result from the subsequent enzyme rearrangements during translocation (subsection “Telomerase translocation leads to complete unfolding of G-quadruplex structure”, Discussion).

6) In Figure 5 and 6 the effects of GQ binding ligands are studied and the conclusion is that these compounds do not completely inhibit telomerase in the smFRET assay. Why are the results of the gel-based assay for the F-22G3 and the smFRET experiments not more similar? It appears in Figure 5L that NMM and PhenDC3 almost completely inhibit the telomerase activity, whereas SST16 does not have as strong of an effect (the authors should quantify the relative activities in this experiment and include them as part of the figure). In contrast, all three compounds yield similar results in the smFRET assay as shown in the bar plots in Figure 5J. The authors should explain this apparent discrepancy, as this might seem to imply that the two-step FRET drop is not correlating with catalytic activity in this experiment.

Quantitation of the activity assays in Figures 5 and 6 has now been included as part of the figures (new Figure 7—figure supplement 5, and Figure 8D). Upon looking at these numbers, we realized that the gel we used to illustrate the effect of NMM on F-22G3 in the original manuscript was not representative of the mean reduction in activity over multiple experiments, most likely due to the NMM causing artefactual retardation of product migration in that particular gel (see old Figure 5L, left panel), so we have swapped this for a more representative gel, without this artefact (new Figure 7—figure supplement 5A). So the reduction in activity with NMM is not as dramatic as it originally appeared; nevertheless, PhenDC3 does still result in a greater effect on activity than on smFRET unwinding. The reason for this is likely that the smFRET data reports only on the first cycle of telomerase-mediated addition of three nucleotides to a single substrate molecule, whereas the ensemble assay is a multiple turnover assay requiring both continued extension of a single substrate, and rebinding to new substrate molecules in solution. Indeed, PhenDC3 affects telomerase in a manner independent of G-quadruplexes, as evidenced by its inhibition of extension of an 18 nt linear telomeric substrate (Figure 7—figure supplement 5 and de Cian et al., 2007) and a completely non-telomeric substrate (de Cian et al., 2007); this effect on telomerase may include preventing it binding to a single-stranded DNA substrate (de Cian et al., 2007), which would affect total product synthesis in a multiple turnover reaction. We have now included discussion of this point in the manuscript (subsection “Ligand stabilization of intramolecular parallel G4 partially inhibits but does not prevent G4 unwinding by telomerase”).

The authors also comment on the impact of PhenDC3 causing a change in the product profile after four repeats 'most likely resulting from G-quadruplex stabilization within the product DNA'. This result is in close accord with the recent study by Jansson et al., 2019 (not cited) wherein the possibility of GQ folding within the product of human telomerase during catalysis is described.

We apologise for missing this citation, which we should have included; we have now done so. This finding has also recently been independently verified (Patrick et al., 2020), so we cited this later paper also (subsection “Ligand stabilization of intermolecular parallel G4 does not inhibit telomerase activity or telomerase-mediated G4 unfolding”).

In addition, in the Discussion, the authors relate the thermal stabilization of the structure by these compounds as evidence for 'telomerase not simply exploiting inherent instability in the G4 structure…' However, smFRET experiments by Jena, Ha et al. JACS 2009 (not cited) – showed that compounds that increase the thermal stability of the GQ fold do not necessarily restrict the conformational dynamics of the structure.

The difference between our study and that of Jena, Ha et al. JACS 2009 is that they used an unmodified human telomeric DNA sequence that adopts a mixture of different G4 topologies that experienced dynamic interconversion between different conformations even in the presence of ligand binding. In our case, smFRET showed no dynamic conversion or unfolding of either of the two G4s we used, with or without ligand binding (compare Figure 2A in their study with Figure 7—figure supplement 3 and Figure 9—figure supplement 1 in our manuscript). Nevertheless, we take the point that thermal stability does not always equal an absence of dynamics, so since we were unable to directly demonstrate a reduction in unfolding dynamics in the presence of ligand (because there was no dynamic unfolding detectable in the time span of our assays in the first place), we have modified that sentence to replace the phrase “provides further evidence that telomerase is not simply exploiting inherent instability in the G4 structure to passively extend spontaneously unwound DNA, but is instead actively involved in G4 unwinding” with “provides evidence that thermal stabilization does not provide a barrier to telomerase-mediated unwinding”.

7) While the two-step FRET signal observed throughout the study is very interesting, the model put forth in Figure 7 is not very satisfying. Throughout the paper (including in the title) the authors describe how these data lead them to a mechanism for how telomerase acts as a GQ resolvase. However, the authors do not make an effort to discuss their data within a structural context based upon the medium-resolution cryo-EM structure of human telomerase (Nguyen et al., 2018, not cited) or the higher-resolution cryo-EM structure of Tetrahymena telomerase (Jiang et al., 2018, not cited). Why is there no attempt to explain how the proposed GQ unfolding might occur? For example, specifically how steric hindrance by the enzyme would cause unfolding during catalysis, but still allow binding to occur at the start of catalysis? Moreover, in the case of the tetrameric GQ, why would activity cause the structure to disassemble as would be suggested by the cartoon in Figure 7 (see comment above)? The authors should clarify this in the text.

We thank the reviewers for this helpful suggestion. In response, we have now enlisted the help of a collaborator with expertise in molecular modelling, Dr Jessica Holien from RMIT University in Australia, to help us determine the feasibility of our proposed model in a structural sense.

Dr Holien had previously constructed an hTERT homology model based on the Tribolium castaneum TERT crystal structure (Tomlinson et al., 2015). She has now refined this model based on the Tetrahymena cryo-EM structure (Jiang et al., 2018). The human cryo-EM structure (Nguyen et al., 2018) was too low a resolution to improve this homology model further, but was used to ensure the resulting model fit in the cryo-EM envelope.

Dr Holien then took the NMR structure of the [7GGT]₄ quadruplex (Gavathiotis and Searle, 2003), and flipped one strand out to allow it to interact with the 10 nt RNA template. She then used the position of an RNA-DNA hybrid in the Tetrahymena cryo-EM structure to position this G4RNA into the active site, with one DNA 3’ end at the catalytic aspartates.

The affinity of telomerase for nucleotides is decreased in the presence of a G4 substrate relative to a linear substrate (Oganesian et al., 2007, Moye et al., 2015), suggesting that the conformation of the active site is perturbed when bound to the bulkier G4. So as we expected, the active site in the modelled enzyme was not large enough to accommodate the G4-RNA complex; it clashed with the C-terminal extension (CTE) of hTERT. However, others have reported that the CTE may pivot around a region called the “beta linker” (Yang and Lee, 2015, Wu, Tam and Collins, 2017; thus, it seemed reasonable that this linker may lengthen to accommodate the quadruplex. The torsions of the beta linker amino acids (Arg938 and Thr939) were therefore altered to enable an extended conformation (shown in Figure 10A with the G4-RNA complex). Note that Figure 10A represents the completely extended form of the CTE; it is likely that during initial binding and extension of the G-quadruplex, the CTE would be less extended than this, and the fully-extended conformation may be needed to allow translocation, as postulated in the papers cited above.

While we emphasize that this model is extremely speculative, we have nevertheless included it in the revised manuscript, since it may help explain our observations. As described above in response to point 2, molecular dynamics simulations suggested that the 5’ ends of the G4 may move slightly apart from each other upon binding in the active site, while the overall structure of the G4 is retained within the boundaries of the active site. Upon DNA-RNA translocation, opening of the active site through pivoting of the CTE domain would allow dissociation of the other 3 DNA strands, resulting in the second step of FRET reduction. We have amended the discussion of our model, and the schematic in Figure 7 (now Figure 10A; described in the subsection “Ligand stabilization of intermolecular parallel G4 does not inhibit telomerase activity or telomerase-mediated G4 unfolding” and in the Discussion), accordingly.

8) The dwell time distribution shown in Figure 1N does not seem to be described particularly well by a single exponential decay. Nonetheless the authors argue that the distribution indicates only one rate-limiting step. Given that the observed reaction is a multi-step process, a more complex kinetic model might be required.

We agree that the dwell time distributions did not fit well with our data and were not very conclusive. According to the reviewer’s suggestion we re-fit our data to a more realistic kinetic model using a gamma function that identifies the number of hidden steps in a complex reaction, if present. For this purpose, we constructed histograms that show distributions of dwell times of inter- and intramolecular G4 (Figure 2—figure supplement 4, Figure 4—figure supplement 4), instead of constructing integrated decay histograms. The new curves show that the data deviate from Gaussian distributions and are more skewed towards the right.

The gamma distribution not only reports on the apparent rate of a complex reaction pathway but also gives the number of rate-limiting steps hidden in the process (Floyd, Harrison and van Oijen, 2010). Both dwell-time histograms were fit to a gamma distribution that gives a number of steps (N) greater than two, suggesting that the process consists of at least two rate-limiting steps. Curve fitting was validated by obtaining chi-square values for the fits; the associated p-values all demonstrate no significant difference between expected and observed values (Figure 2—figure supplement 2, Figure 4—figure supplement 4). These new fits have been introduced into the manuscript, together with a discussion of their meaning.

df: degree of freedom

If the authors have data recorded at various different dNTP concentrations, they should be included and discussed.

Yes, we performed telomerase extension experiments at a 100-fold lower dGTP concentration (5 µM). The data reveal that the G4 unfolding rate decreased by 1.8-fold with the same number of steps (N = 2.1) in comparison to a higher dGTP (0.5 mM) concentration (new Figure 4—figure supplement 4C). This suggests that the rate of G4 unfolding decreased because the telomerase synthesis process was affected by a change in dGTP concentration. This result is discussed in the revised manuscript (subsection “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”).

Related to this point, the second FRET decrease observed with different G4 structures, appears to occur abruptly/instantaneously. Can intermediates during the transition be visualized (at higher frame rates or limiting dNTP concentrations)? If the transition times can be quantified, what is the underlying distribution?

We agree, the transitions seem instantaneous at higher dNTP concentrations. In the experiments at a lower dGTP concentration, we observed that 40 ± 7% molecules of the molecules that changed FRET did indeed display a plateau at 0.4 followed by a gradual drop in FRET from 0.4 to 0.2, compared to a sharp transition at a higher dGTP concentration (new Figure 4—figure supplement 4B). These transitions are gradual and therefore it is challenging to precisely define their endpoints using a change point algorithm. Therefore, we manually analysed those traces to approximate the time it takes to reach 0.2 from 0.4 (which we refer to as τ_unfolding-2, to distinguish it from the dwell time at 0.4), binning those times into a histogram. The histogram was well fitted with a gamma distribution (Chi-square 4.33 and n = 28) that results in ‘N’ greater than 3 with an overall rate constant of 0.44 s^-1 (new Figure 4—figure supplement 4D). This result suggests that at least three hidden steps are present in the transition from 0.4 to 0.2 FRET states, in addition to the 2-3 steps involved prior to this, and that the kinetics of these steps is affected by dNTP concentration, consistent with a dependence on dNTP incorporation. These data are now added and discussed in the revised manuscript (subsection “Human telomerase also binds, unfolds and extends intermolecular parallel G-quadruplexes”). It should be noted that our data are not sufficient to allow us to assign particular kinetic processes to each of these steps, but we do briefly mention some of the steps that may be involved (e.g. nucleotide binding and incorporation, DNA-RNA repositioning, and an enzyme conformational change to reposition the template back in the active site (Parks and Stone, 2014), Discussion).

9) Have EMSAs or similar experiments been performed to clarify that the RNA is actually binding / invading the G4, whereas nothing happens with DNA? Similarly, it would be informative to see binding data (Kd) for telomerase to the G4 structures used here.

We attempted to observe the RNA oligo binding to the G4 by EMSA analysis, but unfortunately this was not successful; we hypothesize that the RNA is too loosely bound to the G4 for the complex to survive electrophoresis (discussed further in response to point 10 below).

The K_m for [7GGT]₄ binding to telomerase was provided in our previous paper (~500 nM; Moye et al., 2015). We have not measured K_d for either G4, since this information is not relevant in the smFRET experiments, in which the DNA is immobilized to a slide.

Also, is the initial FRET change from 0.5 to 0.3 correlated in time with telomerase binding? In other words, does the time lag from injection to FRET shift match the telomerase on-rate, or does the FRET change require more time (indicating a more complex process)?

There is a significant experimental time lag between injection of telomerase and the FRET change, as discussed above. Due to the nonlinear nature of flow in the microfluidic flow cells and the inability to compare injection times from experiment to experiment, it is not possible to quantitatively interpret the initial time before the first change in FRET.

10) Is the FRET change observed by telomerase / RNA binding reversible, e.g. if telomerase dissociates? Similarly, given that very high concentrations of RNA were required to observe the effect seen in Figure 3F, RNA is most likely only loosely bound. It would be interesting to see if the G4 reverts when the RNA dissociates.

No, telomerase enzyme binding to either G4 is not reversible within our experimental window (3-5 min). No association and dissociation kinetics were observed under our experimental conditions.

However, we did observe dynamic association and dissociation of RNA oligos with G4. We performed experiments at RNA oligo concentrations of 100 μm and 500 uM. In those experiments, we categorised static versus dynamic molecules based on whether a given molecule underwent multiple binding and unbinding events (dynamic) or remained at the binding state once bound (static). We observed that 26 ± 6 and 21 ± 4% molecules show dynamic behaviour in 100 µM and 500 µM RNA concentration, respectively, as shown in the FRET time trajectories (Figure 5—figure supplement 2C and D). As the reviewer predicted, this is indicative that the RNA affinity for G4 is much lower than that of the whole enzyme; we have added these data and discussion of this point to the manuscript (new Figure 5—figure supplement 2, and subsection “The RNA template sequence is involved in partial unfolding of G-quadruplex structure”).

11) The authors find that G4-stabilizing ligands double the number of G4 that are not opened. This is not really discussed.

This is a good point, since this happens in the context of there being no increase in the number of molecules undergoing a single FRET drop. This suggests that the ligands may interfere with telomerase binding to F-22G3, preventing the first FRET decrease that is associated with binding. We have added discussion of this point to the manuscript (subsection “Ligand stabilization of intramolecular parallel G4 partially inhibits but does not prevent G4 unwinding by telomerase”).

Could it be that the G4s that are actually unwound are not really bound by the ligands? Does this effect depend on the ligand concentration?

This is possible, although we consider it unlikely since the degree of thermal stabilization of the population of molecules was so dramatic, particularly for PhenDC3 (see Figure 7—figure supplement 2F). We did not conduct experiments at different concentration of ligands, but we carried out ligand binding at an excess of ligand over DNA. Furthermore, at least in the case of NMM, the rate of unwinding was reduced in the presence of ligand, arguing that most molecules were bound by ligand. Nevertheless, we have added to the text the caveat that we cannot rule out that some of the molecules unwound by telomerase were not bound by the ligands (subsection “Ligand stabilization of intramolecular parallel G4 partially inhibits but does not prevent G4 unwinding by telomerase”).

There is also an apparent discrepancy between the amount of G4s that are stabilized, the change in unfolding rate (Figure 5K), and the extension rate (Figure 5L). For example, PhenDC3 results in a minor slowing down of G4 unfolding in Figure 5K but seems to completely inhibit extension (Figure 5L). These discrepancies should be discussed.

We have now re-fitted the dwell time distribution with a gamma distribution function according to the reviewer’s suggestion under question 8. Overall, the unfolding rates slightly decrease in the presence of ligands, though this was only statistically significant in the case of NMM (Figure K).

The difference between the effects of PhenDC3 in the smFRET analyses and the ensemble activity assays have been discussed in response to point 6 above, and this discussion has been added to the revised manuscript (subsection “Ligand stabilization of intramolecular parallel G4 partially inhibits but does not prevent G4 unwinding by telomerase”).

Author details

1. Bishnu P Paudel

   1. Molecular Horizons and School of Chemistry and Molecular Bioscience, University of Wollongong, Wollongong, Australia
   2. Illawara Health and Medical Research Institute, Wollongong, Australia

   Contribution

   Conceptualization, Software, Formal analysis, Investigation, Visualization, Methodology, Writing - original draft, Designed smFRET experiments, performed and analysed smFRET experiments, wrote the manuscript

   Contributed equally with

   Aaron Lavel Moye

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3518-3882

2. Aaron Lavel Moye

   Children’s Medical Research Institute, University of Sydney, Westmead, Australia

   Present address

   Stem Cell Program and Divisions of Hematology/Oncology and Pulmonary & Respiratory Diseases, Boston Children’s Hospital, Harvard Stem Cell Institute, and Department of Genetics, Harvard Medical School, Boston, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Visualization, Writing - review and editing, Designed experiments, purified telomerase and performed ensemble telomerase assays

   Contributed equally with

   Bishnu P Paudel

   Competing interests

   No competing interests declared

3. Hala Abou Assi

   Department of Chemistry, McGill University, Montreal, Canada

   Contribution

   Resources, Formal analysis, Investigation, Writing - review and editing, Synthesis and purification of modified oligonucleotides

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1087-8805

4. Roberto El-Khoury

   Department of Chemistry, McGill University, Montreal, Canada

   Contribution

   Resources, Formal analysis, Investigation, Writing - review and editing, Synthesis and purification of modified oligonucleotides

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8068-4566

5. Scott B Cohen

   Children’s Medical Research Institute, University of Sydney, Westmead, Australia

   Contribution

   Resources, Investigation, Methodology, Writing - review and editing, Purified telomerase and performed ensemble telomerase assays

   Competing interests

   No competing interests declared

6. Jessica K Holien

   School of Science, College of Science, Engineering and Health, RMIT University, Melbourne, Australia

   Contribution

   Formal analysis, Investigation, Writing - review and editing, Performed molecular modeling

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8735-2871

7. Monica L Birrento

   1. Molecular Horizons and School of Chemistry and Molecular Bioscience, University of Wollongong, Wollongong, Australia
   2. Illawara Health and Medical Research Institute, Wollongong, Australia

   Contribution

   Resources, Investigation, Writing - review and editing, Characterization of G4 ligands

   Competing interests

   No competing interests declared

8. Siritron Samosorn

   Department of Chemistry and Center of Excellence for Innovation in Chemistry, Faculty of Science, Srinakharinwirot University, Bangkok, Thailand

   Contribution

   Resources, Supervision, Funding acquisition, Methodology, Writing - review and editing, Synthesis and characterization of G4 ligands

   Competing interests

   No competing interests declared

9. Kamthorn Intharapichai

   Department of Biobased Materials Science, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto, Japan

   Contribution

   Investigation, Writing - review and editing, Synthesis and characterization of G4 ligands

   Competing interests

   No competing interests declared

10. Christopher G Tomlinson

    Children’s Medical Research Institute, University of Sydney, Westmead, Australia

    Contribution

    Supervision, Methodology, Writing - review and editing, Analysis of ensemble telomerase assays

    Competing interests

    No competing interests declared

11. Marie-Paule Teulade-Fichou

    1. Institut Curie, PSL Research University, Orsay, France
    2. Université Paris Sud, Université Paris-Saclay, Orsay, France

    Contribution

    Resources, Supervision, Methodology, Writing - review and editing, Synthesis and characterization of G4 ligands

    Competing interests

    No competing interests declared

12. Carlos González

    Instituto de Química Física ‘Rocasolano’, CSIC, Madrid, Spain

    Contribution

    Resources, Supervision, Methodology, Writing - review and editing, Synthesis and purification of modified oligonucleotides

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-8796-1282

13. Jennifer L Beck

    1. Molecular Horizons and School of Chemistry and Molecular Bioscience, University of Wollongong, Wollongong, Australia
    2. Illawara Health and Medical Research Institute, Wollongong, Australia

    Contribution

    Resources, Supervision, Funding acquisition, Writing - review and editing

    Competing interests

    No competing interests declared

14. Masad J Damha

    Department of Chemistry, McGill University, Montreal, Canada

    Contribution

    Conceptualization, Resources, Supervision, Funding acquisition, Methodology, Writing - review and editing, Synthesis and purification of modified oligonucleotides

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-4458-1623

15. Antoine M van Oijen

    1. Molecular Horizons and School of Chemistry and Molecular Bioscience, University of Wollongong, Wollongong, Australia
    2. Illawara Health and Medical Research Institute, Wollongong, Australia

    Contribution

    Conceptualization, Resources, Supervision, Funding acquisition, Methodology, Writing - review and editing, Designed and analysed smFRET experiments, edited the manuscript

    For correspondence

    vanoijen@uow.edu.au

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-1794-5161

16. Tracy M Bryan

    Children’s Medical Research Institute, University of Sydney, Westmead, Australia

    Contribution

    Conceptualization, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing, Purified telomerase and performed ensemble telomerase assays, analysed data, wrote the manuscript

    For correspondence

    tbryan@cmri.org.au

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-7990-5501

• 2,996

  Page views

• 455

  Downloads

• 25

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.