The 23 bp TerB sequence in interaction with the Tus protein have been successfully used in yeast and mice to create artificial RFBs (Larsen et al., 2014a; Willis et al., 2018; Willis et al., 2017; Willis et al., 2014; Willis and Scully, 2016). To study the effect of site-specific replication blocks at a genomic locus in human cells, we used a previously established MCF7 clone with a unique integrated copy of the pWB15 plasmid carrying two sets of five TerB repeat sequences in opposite, non-permissive orientations with respect to the incoming replication forks (Figure 1A and Figure 1—figure supplement 1). Whole-genome sequencing of the clone was carried out to confirm the single-copy integration of the plasmid at hg38 chr12:95325001. Firstly, we checked whether Tus protein was able to bind TerB sequences in this artificial system in human cells. Chromatin immunoprecipitation (ChIP) followed by quantitative polymerase chain reaction (qPCR) was performed using an antibody against the His-tag 24 hr after MCF7 5C-TerB cells were transfected with a Tus-His expression plasmid. The Tus signal is specifically enriched over 20 times compared to the vector control (VC) condition in proximity to the TerB sequences (PP9, 2, and 47) (Figure 1B).

To corroborate this observation, we performed the proximity ligation assay (PLA) (see ‘Methods’). Two tagged-Tus (HA-Tus and Tus-His) or a VC expression plasmid were simultaneously transfected and antibodies against the two tags, HA and His, were used to generate a PLA signal (Figure 1C and D). We observed 19% of the cells transfected with the two tagged Tus were harboring one or two PLA signal versus 9% of the cells transfected with the VC (Figure 1E), a significant enrichment of the PLA signal when Tus was bound to TerB. These results indicate a specific interaction of the Tus protein with the TerB sequences in human cells.

It has previously been demonstrated that the Tus-TerB interaction can induce an RFB in mouse embryonic cells using a chromosomally integrated system (Willis et al., 2014). We reasoned that our integrated system in MCF7 cells would be more prone to replication fork arrests near the TerB sequences. We used single-molecule analysis of replicated DNA (SMARD) to investigate evidence of replication fork stalling in the genomic DNA (Norio and Schildkraut, 2001). The self-labeling SNAP-tag was used to label Tus. After induction of Myc-NLS-TUS-SNAP by using Dox for 3 d, asynchronous MCF7 5C-TerB cells were sequentially pulse-labeled using two nucleoside analogs (iododeoxyuridine [IdU] and chlorodeoxyuridine [CldU]) for 4 hr each (Figure 2A). The isolated DNA was digested by restriction endonuclease Sfil, and the 200 kb fragment obtained was stretched on slides for staining. Two DNA FISH probes were used to detect the 200 kb segment of interest containing the TerB arrays: one probe of 40 kb within the chromosome 12 (hg38 chr12:95,199,827–95,239,226) and one probe of 7 kb made from pWB15 (Figure 2B). The direction of replication forks could be determined using the labeling patterns and its position within the locus of interest using our FISH probes. After inducing the expression of Myc-NLS-TUS-SNAP from an integrated plasmid in MCF7 5C-TerB cells (Figure 2—figure supplement 1A and B), we analyzed its impact on the replication fork progression within the region containing TerB repeats (Figure 2C and D). The yellow arrows indicate the transition of labeling from IdU to CldU incorporation showing the replication fork direction. When Tus was expressed, these transition sites were found accumulated near the TerB regions (Figure 2D, arrows within the white oval) compared to the random signal spread on the 200 kb region when Tus is not expressed (Figure 2C). We quantified the percentage of molecules with replication forks at each 5 kb interval in the 200 kb SfiI segment containing TerB sequences (Figure 2E and F) and confirmed an accumulation of forks at the TerB sequences only when Tus was expressed (Figure 2F, black arrow), supporting that the binding of Tus on TerB arrays can impair the replisome progression at genomic loci. Interestingly, expression of the Tus protein alone does not alter the direction of replication fork progression along the 200 kb SfiI segment where forks predominantly progress from the 3′ to 5′ direction in cells both with and without Tus expression as analyzed by the percentage of molecules with IdU incorporation at each 5 kb interval (Figure 2—figure supplement 1C and D). Furthermore, no significant difference was observed in the global replication fork speed in cells with or without Tus expression (Supplementary file 1a). Together, these data indicate that when Tus bound the TerB sequences integrated within chromosome 12, it creates a physical barrier and an obstacle to the replication fork progression.

The eukaryotic replisome is a multiprotein apparatus consisting of DNA helicases, DNA polymerases and accessory proteins to duplicate the genome. We hypothesized that an RFB would induce an accumulation of these proteins on chromatin around TerB sequences. To test our hypothesis, we performed ChIP using an antibody against mini-chromosome maintenance protein 3 (MCM3), a CMG core-complex protein, in the absence or presence of Tus. When Tus was expressed, MCM3 was threefold more enriched upstream of the TerB arrays at the PP0-2 and PP10 sites compared to the surrounding primer pairs (Figure 2G and H). Additionally, we find that there is an enrichment of the DNA repair scaffold protein FANCM at the TerB sites mirroring the MCM3 enrichment (Figure 2—figure supplement 2), indicative of the stalled replication forks as previously described (Collis et al., 2008; Nandi and Whitby, 2012; Panday et al., 2021; Xue et al., 2015). Together, these results show that Tus expression induces an RFB at the TerB arrays impairing the replisome progression during S-phase.

One of the earliest responses to replication stress is the phosphorylation of the histone variant H2AX on serine 139 (γH2AX) by members of the phosphoinositide 3-kinase (PI3K)-like family (PIKK) (Ward and Chen, 2001). As we demonstrated that TerB arrays can block incoming replication forks, we asked whether we could see a γH2AX signal enrichment due to fork arrest. Using a co-immunoprecipitation assay with the chromatin fraction of cross-linked cells expressing either GFP tag or GFP-Tus, we found that GFP-Tus and γH2AX were immunoprecipitated together using GFP or γH2AX antibodies (Figure 3A).

To gain insight into the γH2AX signal at the RFB induced by the Tus-TerB interaction, cells were transfected with either VC or HA-Tus expression plasmids, and PLA was performed using antibodies against HA and γH2AX (Figure 3B). We found that 16% of the HA-Tus transfected cells were harboring at least one PLA signal versus 3% of the VC transfected cells (Figure 3C and D).

To better characterize the γH2AX enrichment around TerB sites, we performed ChIP-qPCR assays using a γH2AX antibody 24 hr after Tus or VC expression. We observed an enrichment of γH2AX, strictly co-localizing with the Tus-Ter interaction sites (PP9 and PP47), suggesting a local role of γH2AX in response to stalled RFs (Figure 3E and F). This is in stark contrast to the observed spreading of the γH2AX signal after a site-directed DSB (Berkovich et al., 2007; Chailleux et al., 2014; Clouaire et al., 2018; Savic et al., 2009). To corroborate this observation in our integrated cassette, we generated a site-specific DSB by expressing Cas9 targeted to the integrated cassette (Figure 3—figure supplement 1A–C). We assessed γH2AX using the ChIP-qPCR assay 24 hr after transfection and noticed enrichment at the distal primer pairs (PP0-2 and PP10), contrasting with the very tight local signal observed at the Tus-TerB RFB (Figure 3—figure supplement 1D). However, we did not find any significant γH2AX signal between VC or Tus expression (Figure 3—figure supplement 2A and B), supporting the lack of global response. This suggests that the RFB generated by the interaction between Tus and the TerB arrays activates a stress response that stimulates the phosphorylation of H2AX, concentrated around the fork block.

ATR is known to phosphorylate H2AX in response to replication stress and plays a pivotal role in the surveillance of DNA replication (Ward and Chen, 2001). To understand whether the global ATR-dependent checkpoint was activated in response to the Tus-TerB RFB, we analyzed the levels of phospho-proteins involved in the intra-S checkpoint activation pathway (RPA, CHK1, and ATR) by immunoblotting fractionated cell extracts 24 hr after VC or Tus expression. Cells treated with 2 mM hydroxyurea (HU) for 4 hr demonstrated an increase in phospho-protein levels, acting as a control for checkpoint activation (Figure 4A, lanes 3 and 4). Tus expression alone did not activate the global ATR-dependent checkpoint as the level of phospho-protein remained the same as the VC control (Figure 4A, lanes 1 and 2). We did not find any observable difference in cell cycle progression between VC or Tus expression alone (Figure 4—figure supplement 1A and B), supporting the lack of global response.

We hypothesized that if the local γH2AX enrichment at TerB sites was ATR-dependent, a reduction in ATR activity would lead to a decreased γH2AX signal. To validate the ATR inhibitor activity (ATRi), VE-822 (Fokas et al., 2012), we performed immunoblotting using total protein extracts of cells treated with 2 mM HU and compared the level of phospho-ATR TH1989 with or without an ATR inhibitor. A 4 hr treatment with ATRi reduced the level of phospho-ATR TH1989, reflecting a decrease in ATR activity (Figure 4B). ChIP-qPCR assays were performed using a γH2AX antibody 24 hr after VC or Tus expression, and cells were harvested after a 4 hr ATRi treatment. Consistent with our hypothesis, the γH2AX enrichment in the vicinity of TerB sites after Tus expression was remarkably decreased upon ATRi treatment (Figure 4C–E, PP9 and PP47), implying that the Tus-TerB RFB activates a localized stress response that stimulated the local phosphorylation of H2AX. Together, these data suggest a local activation of the intra-S checkpoint via the ATR kinase and rapid phosphorylation of H2AX, which could mediate the recruitment of repair factors near the damage site.

MCF7 5C-TerB cells were grown at 37°C and 5% CO₂ in complete DMEM high glucose supplemented with 10% FBS, 2 mM GlutaMAX, 20 mM HEPES, 100 I.U./ml penicillin, and 100 μg/ml streptomycin supplemented with 400 µg/ml zeocin. The cell line was characterized using the Cell Line Authentication by STR analysis at the Integrated Genomics Operation, MSKCC. Cells were routinely tested for mycoplasma contamination using the Universal Mycoplasma Detection Kit (ATCC 30-1012K).

For PLA, cells were seeded at 1 × 10⁵ in a 12-well cell culture plate. For T7 assay, cells were seeded at 2.5 × 10⁵ in a 6-well cell culture plate. For ChIP, IP, and cellular fractionation, cells were seeded at 2.5 × 10⁵ in a 10 cm cell culture dish.

MCF7 5C-TerB cells were seeded and transfected the day after with 10 μg of pCMV3xnls Tus or pCMV3xnls using the Mirus Bio TransIT-LT1 reagents for 24 hr more. For the generation of DSB, cells were transfected 6 hr prior to harvesting with 120 pmol of the sgRNA TerB1 and 120 pmol of the Cas9 protein (obtained from QB3 Macrolab, Berkeley) using the Lipofectamine CRISPRMAX Cas9 transfection reagents.

PvuI linearized pWB15 was transfected into MCF7 (ATCC HTB-22) cells to generate the stably integrated clone, MCF7 5C-TerB (Figure 1—figure supplement 1). Then, 800 μg/ml of zeocin was used to select positively integrated cells and monoclonal colonies were isolated and subsequently maintained in 400 μg/ml zeocin. Single-integrant validation was performed by whole-genome sequencing of the clone.

Stable MCF7 5C-TerB cells-inducible-expressing Myc-NLS-TUS-SNAP were generated by lentiviral transduction. Cells were infected with lentiviral particles containing pInducer10L or pInd Tus-SNAP, and single-cell clonal colonies were selected in complete DMEM containing 1 µg/ml puromycin and 400 µg/ml zeocin. Myc-NLS-TUS-SNAP protein was expressed in stable lines by induction with 1 µg/ml doxycycline for 3 d.

The doxycycline-inducible lentiviral plasmid used to express N-terminally myc-tagged, C-terminally SNAP-tagged nuclear localized, human codon-optimized wild-type Tus (Myc-NLS-TUS-SNAP) was generated as follows. The N-terminally myc epitope-tagged, nuclear-localized, codon-optimized wild-type Tus cDNA sequence was PCR-amplified from pcDNA3-β-MYC-NLS-Tus using primers forward 5′-AGTCGGTACCGAATTCGCCACCATGGAACAAAAGCTG-3′ and reverse 5′-AGTCGGCGGCCGCGCCGCTACCGTCAGCCACGTACAGGTGCA and the SNAP-tag cDNA sequence PCR amplified using primers forward 5′-AGTCGCGGCCGCCGGCCACATGGACAAAGACTGCGAAATGAAGC-3′ and reverse 5′-ACTGCTCGAGTCAACCCAGCCCAGGCTTGC-3′. The Tus and SNAP PCR-amplified fragments were digested with KpnI and NotI, and NotI and XhoI, respectively, and inserted into the inducible expression lentiviral vector pInducer10L (Drosopoulos et al., 2020) directly downstream of the doxycycline-inducible promoter, to generate pInd Tus-SNAP. This construct was sequenced to confirm that unintended mutations were not introduced during PCR and cloning.

Cells were harvested by trypsinization, suspended in complete DMEM, washed with PBS, then pelleted and flash frozen in liquid N₂ and stored at –80°C. For SDS-PAGE, pellets were thawed on ice and lysed by resuspending in Laemmli buffer (60 mM Tris-HCl pH 6.8, 400 mM 2-mercaptoethanol, 2% SDS, 10% glycerol, 0.01% bromophenol blue) to a final concentration of 10⁶ cells/ml. Lysates were denatured (5 min at 100°C) and passed through a 25-gauge needle (5×) then spun for 2 min at full speed in a microfuge. Aliquots of lysate corresponding to 10⁵ cells were resolved on 4–15% gradient SDS-PAGE gels, proteins transferred to nitrocellulose membrane and blocked in PBS with 5% Blotting-grade Blocker and 0.1% Tween 20. Membranes were then incubated with primary antibodies diluted in PBS with 5% Blotting-grade Blocker. Primary antibodies used were anti-Myc tag and anti-actin. Following incubation with primary antibodies, membranes were washed with PBS + 0.1% Tween 20. Membranes were then incubated with fluorescently labeled goat anti-mouse IRDye 680LT and goat anti-rabbit IRDye800CW secondary antibodies and then washed in PBS + 0.1% Tween 20. Immunoblots were imaged on an Odyssey Lc Infrared scanner.

SMARD was performed essentially as previously described (Drosopoulos et al., 2012; Norio and Schildkraut, 2001; Twayana et al., 2021). Exponentially growing cells in complete DMEM with 1 µg/ml doxycycline were sequentially pulse-labeled with IdU (4 hr), followed by CldU (4 hr) each to a final concentration of 30 μΜ. Following pulsing, cells were suspended in PBS at a concentration of 3 × 10⁷ cells per ml. An equal volume of 1% molten TopVision low-melting point agarose in PBS was added. The resulting cell suspension in 0.5% TopVision agarose was poured into wells of a chilled plastic mold to make plugs of size 0.5 cm × 0.2 cm × 0.9 cm, each containing 10⁶ cells. Cells in the plug were lysed at 50°C in a buffer containing 1% n-lauroylsarcosine, 0.5 M EDTA, and 0.2 mg/ml proteinase K. Plugs were rinsed in TE and washed with 200 μM phenylmethanesulfonyl fluoride (PMSF). Genomic DNA in the plugs was digested with SfiI overnight at 37°C. Digested genomic DNA was cast into 0.7% SeaPlaque GTG agarose gel, and DNA was separated by pulsed field gel electrophoresis (PFGE) using a CHEF-DRII system (Bio-Rad). The 200 kb DNA segment containing TerB sequences within the gel was located by performing Southern hybridization on a portion of the gel using a specific probe made from the pWB15 plasmid. This identified the pulse field gel slice containing TerB sequences, which was excised and melted (20 min at about 72°C). The DNA in the gel solution was stretched on microscope slides coated with 3-aminopropyltriethoxysilane, denatured with sodium hydroxide in ethanol, fixed with glutaraldehyde, and hybridized overnight with biotinylated DNA FISH probes at 37°C in a humidified chamber. Biotinylated DNA FISH probes were made from the fosmid WI2-1478M20 and from the plasmid pWB15. Following hybridization, slides were blocked with 3% BSA for at least 20 min and incubated with the avidin Alexa Fluor 350. This was followed by two rounds of incubation with the biotinylated anti-avidin D for 20 min, followed by the avidin Alexa Fluor 350 for 20 min. Slides were then incubated with an antibody specific for IdU; a mouse anti-BrdU antibody, an antibody specific for CldU; a rat monoclonal anti-BrdU antibody, and the biotinylated anti-avidin D for 1 hr. This was followed by incubation with secondary antibodies: Alexa Fluor 568 goat anti-mouse IgG (H+L), Alexa Fluor 488 goat anti-rat IgG (H+L), and the avidin Alexa Fluor 350 for 1 hr. Slides were rinsed in PBS with 0.03% IGEPAL CA-630 after each incubation. After the last rinse in PBS/CA-630, coverslips were mounted on the slides with ProLong Gold Antifade reagent. A Zeiss fluorescent microscope and IP Lab software (Scanalytics) were used to capture images of IdU/CldU incorporated DNA molecules. Images were processed using Adobe Photoshop and aligned (using Adobe Illustrator) based on the positions of FISH signals that identify the 200 kb segments that contain the ter sequences.

MCF7 5C-TerB cells were seeded on poly-L-lysine-coated coverslips and transfected the day after as described before. Then, 24 hr after transfection, cells were washed with PBS, and fixed and permeabilized with 4% paraformaldehyde containing 0.2% Triton X-100 for 20 min on ice and blocked with PBS-BSA 3% overnight at 4°C. Coverslips were incubated with primary antibodies for 1 hr at room temperature (RT). Proximity ligation was performed using the Duolink In Situ Red Starter Kit Mouse/Rabbit (Sigma-Aldrich) according to the manufacturer’s protocol. The oligonucleotides and antibody-nucleic acid conjugates used were those provided in the Sigma-Aldrich PLA kit. Images were quantified by counting the number of foci per nucleus using Nikon software.

For immunoprecipitation, MCF7 5C-TerB cells were pre-extracted using CSK100 buffer (100 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 10 mM PIPES pH 6.8, 1 mM EGTA, 0.2% Triton X-100, anti-protease and anti-phosphatase) for 5 min on ice, fixed in 1% formaldehyde for 10 min on ice, and a 1% glycine solution was used to stop the reaction. After scraping the cells in ice-cold PBS and centrifugation, pellets were lysed in SDS lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% SDS, anti-proteases, and anti-phosphatases) for 10 min on ice and sheared for 3 min. Samples were cleared by centrifugation for 5 min at 4°C. Immunoprecipitations were performed on 10-fold diluted lysates in dilution buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.2% Triton X-100, anti-proteases and anti-phosphatases) with antibodies against GFP, γH2AX, or IgG overnight at 4°C on a wheel. Beads were extensively washed in the dilution buffer and denatured in 2× Laemmli buffer.

Proteins were separated on 4–12% acrylamide SDS-PAGE, transferred on nitrocellulose membrane, and detected with the indicated antibodies described in Supplementary file 1b and ECL reagents.

For the cellular fractionation, MCF7 5C-TerB cells were scraped in PBS, divided into two different tubes (1/3 of the volume for the whole-cell extract [WCE] and the remaining 2/3 for the fractionation) and centrifuged to keep the pellets.

For the WCE, cells were lysed in 1 volume of lysis buffer (50 mM Tris pH 7.5, 20 mM NaCl, 1 mM MgCl₂, 0.1% SDS, anti-protease and anti-phosphatase) for 10 min at RT on a wheel and denaturated in 2× Laemmli buffer.

For the fractionation, MCF7 5C-TerB cells were pre-extracted in 2 volumes of CSK100 (100 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 10 mM PIPES pH 6.8, 1 mM EGTA, 0.2% Triton X-100, anti-protease and anti-phosphatase) for 15 min on ice, and centrifuged. The supernatant (SN) representing the soluble fraction was kept in a new tube (soluble fraction). The pellet was washed with CSK50 (50 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 10 mM PIPES pH 6.8, 1 mM EGTA, 0.2% Triton X-100, anti-protease and anti-phosphatase) and resuspended in 2 volumes of CSK50 containing benzonase for 10 min on a rotating wheel, and the SN was kept after centrifugation (chromatin fraction). All the fractions were denaturated in 2× Laemmli buffer.

Proteins were separated on 4–12% acrylamide SDS-PAGE, transferred on nitrocellulose membrane, and detected with the indicated antibodies described in Supplementary file 1b and ECL reagents.

ChIPs were performed using the ChIP-IT express according to the manufacturer’s instructions. Cells were crosslinked, washed, harvested by scraping, and lysed according to the kit’s instructions. Chromatin was sonicated to 200–1500 bp fragments and DNA concentration determined. Equal amount up to 15 ug of DNA was then used for each immunoprecipitation. The antibodies used are listed in Supplementary file 1b. DNA fragments were eluted, purified, and analyzed by SYBR Green real-time PCR. The sequences of primers used for qPCR are given in Supplementary file 1b. Experiments were repeated at least three times, and each real-time qPCR reaction was performed in duplicate.

MCF7 5C-TerB cells were seeded and transfected the day after with 120 pmol of Cas9 protein and 120 pmol of the sgRNA TerB1 using the Lipofectamine CRISPRMAX Cas9 transfection reagents (previously described). Then, 24 hr post-transfection, cells were lysed using a lysis buffer (100 mM NaCl, 10 mM Tris-HCl pH 8, 25 mM EDTA pH 8, 0,5% SDS) and 50 ug of Proteinase K overnight at 50°C with shaking. After addition of NaCl and mix for 1 min, the supernatant was put in a new tube and an ethanol precipitation was performed. A PCR to amplify the 900 bp region surrounding the sgRNA TerB1 site was conducted using PP9F and R. Heteroduplex were formed by heating and cooling down the samples, and a T7 endonuclease assay digestion was performed prior to electrophoresis for detection.

MCF7 5C-TerB cells were transfected as described before. Control cells were treated with 2 mM HU for 4 hr before harvesting. Cells were washed with PBS and fixed in 1 ml cold 70% ethanol for 30 min. Cells were pelleted and washed with PBS. Then, 100 µg/ml RNase A was added and 50 µg/ml PI solution was added directly to the pellet, mixed well, and incubated for 30 min at RT in the dark. A total of 50,000 cells per condition were analyzed by flow cytometry.

Cells were fixed with 4% PFA/PBS for 20  min, permeabilized with 0.5% Triton-X 100 for 10  min, washed three times in 1× PBS, and blocked in 10% goat serum overnight at 4°C. The primary antibody (1:500) was incubated for 2 hr at RT. Cells were then washed three times in 1× PBS. The secondary antibody (1:1500) was incubated for 1 hr at RT in dark. Cells were then washed again three times in 1× PBS. Stained cells were mounted with mounting medium containing DAPI. Slides were imaged at 60× (immersion oil) using Nikon A1 spinning disk confocal microscope.

For PLA and gH2AX experiments, slides were imaged at 60× (immersion oil) with Nikon spinning disk confocal microscope. PLA foci per nucleus and gH2AX foci per nucleus were calculated using Nikon Elements AR Analysis Explorer (version 5.21.03), where DAPI was used as a mask for the nucleus. The number of PLA foci per nucleus was quantified to get the percentage of cells with one or two foci indicating a positive signal at the site-specific replication block. The number of gH2AX foci was counted for each DAPI to obtain the average number of gH2AX foci in each condition.

Genomic DNA was extracted from the cells and submitted to Novogene for sequencing. The samples were processed and whole-genome sequencing was performed using their hWGS service pipeline.

Statistical analysis for the various experiments was performed using GraphPad Prism version 9.4.0. Results are presented as mean ± standard error of the mean (SEM). A p value of <0.05 by Student’s t-test was considered statistically significant. ns, nonsignificant, *p<0.05, **p<0.005, ***p<0.001.