TRF1 relies on fork reversal to prevent fragility at human telomeres

Abstract

Telomeres pose challenges during replication, with converging forks unlikely to resolve issues. Depleting TRF1 results in fragile telomeres, yet its exact role in telomere replication remains unclear. In our cellular model, insufficient TRF1 density at long telomeres leads to telomere fragility that is alleviated by restoring telomeric TRF1 levels. Our findings indicate that TRF1 mitigates lagging strand telomere fragility through fork reversal in a process involving telomerase activity, rather than merely alleviating fork barriers. Additionally, TFIIH, a crucial partner of TRF1, aids in restarting replication on the leading strand after fork reversal. When fork reversal is compromised, PrimPol-mediated repriming rescues fragility at leading strand telomeres, revealing a new role for this enzyme in human telomere replication. Lastly, our findings indicate that the TRF1-mediated decrease in telomere fragility is dependent on RNA:DNA hybrids, likely facilitating fork restart.

Fig. 1. Telomere fragility is associated with telomere length.

A Telomerase activity assessed by ddTRAP in PCS clones, normalized to PCS-8. SW39/TEL⁺ and IMRB/ALT⁺ cells (grey bars) were used as positive and negative controls, respectively. n = 2 independent biological replicates, except for PCS-7 and SW39 (n = 1). Mean ± SEM. Purple: clones with long telomeres; Cyan: clones with short telomeres. B TRF analysis (n = 1), with PCS-2 designated as PCS^LT and PCS-3 designated as PCS^ST. C qRT-PCR analysis of hTERT expression, normalized to ACTB expression and to PCS^ST. n = 3 independent biological replicates, except for PCS^LT cells (n = 4). Mean ± SEM. D Left: Representative FISH pictures on metaphase spreads from PCS^ST and PCS^LT cells showing fragile telomeres (arrowheads). Right: Fragile telomeres with U2OS/ALT⁺ and IMRB/ALT⁺ cells as controls for ALT⁺ cells. A minimum of 5000 chromosome ends were analyzed per condition, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis test. E Representative blot (above) and quantification (below) for C-circle assay (CCA) in PCS^ST and PCS^LT cells, with SW39/TEL⁺ and IMRB/ALT⁺ cells as negative and positive controls, respectively. Signals were normalized to PCS^ST. n = 3 (SW39/TEL⁺ and IMRB/ALT⁺) or n = 4 biological replicates (PCS^ST and PCS^LT). Mean ± SEM. Two-tailed ratio paired t-test. F TRF analysis (above) and CCA blot (below). PD population doubling. n = 1. G Fragile telomeres assessed by FISH on metaphase spreads. A minimum of 2200 chromosome ends were analyzed, based on n metaphase spreads collected from one experiment for PCS^ST-hTERT at PD9 and PD78, two independent biological replicates for PCS^ST-hTERT at PD18-21 and PD86-89, and three independent biological replicates for PCS^ST and PCS^LT. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis test. Source data are provided as a Source Data file.

Fig. 2. PCS^LT cells display a subset of ALT features.

A Fragile telomeres assessed by either FISH (blue) or CO-FISH (black) on metaphase spreads. FISH data are from Fig. 1D. For CO-FISH, fragility was the average fragility from both the leading and lagging strands. U2OS/ALT⁺ and IMRB/ ALT⁺ cells were used as controls for ALT⁺ cells. A minimum of 3800 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Two-tailed unpaired t-tests with Welch’s correction if needed. B Left: Representative CO-FISH pictures on metaphase spreads from PCS^LT cells showing lagging (green) and leading (red) telomere fragility (arrowheads). Right: Lagging (green) and leading (black) telomere fragility. A minimum of 5000 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Two-tailed Mann–Whitney tests. C Representative blots (above) and quantification (below) for CCA in siPOLD3-treated PCS^LT cells, using U2OS/ALT⁺ cells as control. Signals were normalized to siLuc. n = 5 (U2OS/ALT⁺) or n = 6 (PCS^LT) independent biological replicates. Mean ± SEM, Two-tailed ratio paired t-tests. D Left: Representative images for Telo-FISH (red)/IF targeting BLM (cyan) and PML (green) in PCS^LT cells, with U2OS/ALT⁺ cells as positive control. Arrowheads indicate PML-associated telomeres bound by BLM. Right: Percentage of PML-associated telomeres co-localizing with BLM per nucleus. n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Two-tailed Mann–Whitney test. E Above: Representative CO-FISH pictures on metaphase spreads from IMRB/ALT⁺ cells showing T-SCE events (arrowheads). Below: T-SCEs in PCS^ST and PCS^LT, with U2OS/ALT⁺ and IMRB/ALT⁺ as positive controls. n metaphases were analyzed from three independent biological replicates. Median in red, quartiles in orange. Kruskal–Wallis test. F Above: Representative images from native Telo-FISH targeting ssC-rich telomeric DNA (red) in U2OS/ALT⁺ and IMRB/ALT⁺ cells. Below: ssTeloC signals in PCS^ST and PCS^LT, with U2OS/ALT⁺ and IMRB/ALT⁺ cells as positive controls. n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Kruskal–Wallis test. Source data are provided as a Source Data file.

Fig. 3. Replication stress markers are enriched at PCS^LT telomeres.

A Above: Western blot of cell extracts from PCS^ST and PCS^LT cells, probed for pRPA32-S33, RPA32, and Actin. Below: pRPA32-S33 (purple) and RPA32 (green) levels, normalized to Actin and to PCS^ST. Representative of n = 4 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-tests. B Left: Representative images for Telo-FISH (red)/IF experiments targeting RAD51 (cyan) and RPA (green) in PCS^ST and PCS^LT cells. Right: Co-localization events between telomeres and RAD51 (cyan) or RPA (green) in PCS^ST and PCS^LT. n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Two-tailed Mann–Whitney tests. C Cellular viability of PCS^ST and PCS^LT cells treated with siRAD51. 72 h after transfection in 6-wells plate, cells were counted using Burker slide. n measurements from three independent biological replicates and technical replicates for PCS^ST (n = 6) and PCS^LT (n = 13). Mean ± SEM. Two-tailed unpaired t-test. D Representative blot (above) and quantification (below) of CCA in siFANCM-treated PCS^LT cells. Signals were normalized to siLuc. n = 8 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-test. E Lagging (green) and leading (black) telomere fragility assessed by CO-FISH in siFANCM-treated PCS^LT cells. A minimum of 8300 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Two-tailed unpaired t-test with Welch’s correction for the lagging strand comparison and two-tailed Mann–Whitney test for the leading strand comparison. F Left: Alkaline gel electrophoresis of ClaI-digested genomic DNA from PCS^ST and PCS^LT cells. Ethidium bromide (EtBr) staining is shown as loading control. Right: Quantification of alkaline gel signals using the G-rich probe. The radioactive signal intensity below 6 kb was normalized to the total lane signal. The normalized signal from PCS^ST cells was set to 1 for comparison. n = 3 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-test. G 2D-gel analysis with telomeric probe showing i-loops (red arrows). Ethidium bromide staining is shown as loading control. U2OS/ALT⁺ cells were used as positive control. n = 1. Source data are provided as a Source Data file.

Fig. 4. RNaseH1 overexpression exacerbates leading telomere fragility in PCS^LT cells.

A DNA:RNA immunoprecipitation (IP) analysis using S9.6 antibody in extracts from PCS^ST and PCS^LT-EV cells. IP and input samples were analyzed by qPCR with primer sets amplifying telomeric DNA. Values in IP were normalized first to input and then to PCS^LT-EV. Values obtained for control experiments with IgG were subtracted. n = 3 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-test. B Western blot of cell extracts from PCS^LT cells transduced with pLHCX empty vector (EV), pLHCX-RNaseH1-Myc (RH1) or pLHCX-catalytically dead RNaseH1-Myc (RH1-CD), probed for Myc and Actin. Representative of n = 6 biological replicates. C Lagging (green) and leading (black) telomere fragility assessed by CO-FISH in PCS^LT cells transduced with EV, RH1, or RH1-CD. A minimum of 11500 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis tests. D Left: Representative images for Telo-FISH (red)/IF experiments targeting RAD51 (cyan) and RPA (green) in PCS^LT cells transduced with EV, RH1, or RH1-CD. Right: Co-localization events between telomeres and RAD51 (cyan) or RPA (green). n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Kruskal–Wallis tests. Source data are provided as a Source Data file.

Fig. 5. Suboptimal TRF1 binding at PCS^LT telomeres causes telomere fragility.

A Above: Representative blot from ChIP analyses showing TRF1 binding at telomeres in PCS^ST and PCS^LT cells and in PCS^LT cells transduced with pLPC empty vector (EV) or pLPC-TRF1 (TRF1). Below: Quantifications. Signals were normalized first to telomeric input signals and then to either PCS^LT (left part) or EV-transduced PCS^LT cells (right part). n = 3 independent biological replicates for PCS^LT and PCS^ST and n = 4 independent biological replicates for PCS^LT cells transduced with either EV or TRF1. Mean ± SEM. Two-tailed ratio paired t-tests. B Western blot of cell extracts from PCS^LT cells transduced with pLPC/pLHCX-empty vector (EV), pLPC-TRF1 (TRF1) or pLHCX-RNaseH1-Myc (RH1), probed for TRF1 and Myc. Representative of n = 3 biological replicates. C Above: Representative CO-FISH pictures on metaphase spreads from PCS^LT cells transduced with either EV, TRF1, or both TRF1 and RH1 showing lagging (green) and leading (red) telomere fragility (arrowheads). Below: Lagging (green) and leading (black) telomere fragility. A minimum of 5200 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis tests. Source data are provided as a Source Data file.

Fig. 6. TRF1-dependent reduction of telomere fragility is not linked to the loss of replication stress markers.

A Above: Western blot of cell extracts from PCS^LT cells transduced with pLPC empty vector (EV) or pLPC-TRF1 (TRF1), probed for pRPA32-S33, RPA32, and Actin. Below: pRPA32-S33 (purple) and RPA32 (green) levels, normalized to Actin and to PCS^LT-EV. Representative of n = 6 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-tests. B Co-localization events between telomeres and RAD51 (cyan) or RPA32 (green) in PCS^LT cells transduced with either EV or TRF1. n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Two-tailed Mann–Whitney tests. C Representative blot and quantification for CCA in PCS^LT cells transduced with either EV or TRF1. Signals were normalized to PCS^LT-EV. n = 3 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-test. D Co-localization events between telomeres and PML in PCS^LT cells transduced with either EV or TRF1. n nuclei were analyzed from three independent biological replicates. Median in red, quartiles in orange. Two-tailed Mann–Whitney test. E Above left: Schematic representation of the pTelN mini-chromosome containing 115 telomeric repeats (red arrow) and the SV40 replication origin (ORI). Above right: Schematic representation of the 2D-gel migration profile of the specified DNA replication intermediates. Below left: 2D-gel analysis of plasmid DNA using the 5.2 kb BamHI-SacI probe, in cells with and without ectopic overexpression of TRF1. Below right: Quantification of cone signal (reversed fork migration) as a ratio of cone signal to total replication intermediates (RI), normalized to EV. n = 3 independent biological replicates. Mean ± SEM. Two-tailed ratio paired t-test. F, G Lagging (green) and leading (black) telomere fragility assessed by CO-FISH in PCS^LT-EV or PCS^LT-TRF1 cells under the specified conditions. siSM1: siSMARCAL1. Cells were treated with either 15 µM BIBR (F) or 5 µM Olaparib (PARPi) (G), or DMSO as control, during 48 h before BrdU/BrdC incorporation. A minimum of 8600 (F) or 6200 (G) chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis tests. Source data are provided as a Source Data file.

Fig. 7. TRF1’s capacity to reduce telomere fragility depends on fork reversal, PrimPol activity, and RNA-DNA hybrids.

A Lagging (green) and leading (black) telomere fragility assessed by CO-FISH in PCS^LT-EV, PCS^LT-TRF1 or PCS^LT-TRF1-RH1 cells under the specified conditions. siSM1: siSMARCAL1. Cells were treated with 60 µM spironolactone (SP), or DMSO as control, during 25 h before BrdU/BrdC incorporation. A minimum of 5900 chromosome ends were analyzed, based on n metaphase spreads collected from three independent biological replicates. Boxplot: Min, 1st quartile, Median, 3rd quartile, and Max. Kruskal–Wallis tests. Source data are provided as a Source Data file. B Model. Stalled replication forks at PCS^LT telomeres undergo SMARCAL1-dependent reversal, generating new 3′ telomeric ends. The reversed forks are bound by telomerase. Fork replication can resume through a homologous recombination-dependent restart process that requires RAD51 and might be facilitated by long 3′ overhangs generated by the telomerase and the formation of RNA:DNA hybrid formation, possibly involving TFIIH. If reversal is suppressed, PrimPol-dependent repriming can promote fork restart. This process may also be dependent on RNA:DNA hybrids. Optimal levels of TRF1 are required to suppress telomere fragility, by facilitating the reversal/restart processing of stalled forks. Protein illustrations were generated using BioRender (Vaurs, M. (2025) https://BioRender.com/ke04c93).