Selective interactions at pre-replication complexes categorize baseline and dormant origins

Abstract

DNA synthesis in metazoans initiates within a select group of replication origins (baseline origins), whereas other (dormant) origins do not initiate replication despite recruiting apparently indistinguishable pre-replication complexes. Dormant origins are activated as backups when DNA synthesis stalls, allowing for complete genome duplication, yet it is unclear how cells selectively differentiate between baseline and dormant origins. We report here that during unperturbed cell proliferation, dormant origins selectively bind phosphorylated RecQL4 (pRecQL4), a member of the RecQ helicase family mutated in Rothmund-Thomson, RAPADILINO and Baller-Gerold syndromes. Origin-bound pRecQL4 prevents the binding of an essential replication initiation complex, MTBP-TICRR/TRESLIN, to dormant origins, thus restricting replication initiation to baseline origins. When cells encounter replication stress, pRecQL4 is required for the dissociation of the MTBP-TICRR/TRESLIN complex from chromatin, which, in turn, facilitates the subsequent redistribution of MTBP-TICRR/TRESLIN to both baseline and dormant origins and allows recovery from replication inhibition. Thus, the interactions between the MTBP-TICRR/TRESLIN complex and pRecQL4 at replication origins are critical for replication origin choice and facilitate recovery from replication stress.

Fig. 1. Association of MTBP and phospho-RecQL4 with pre-RCs at baseline and dormant replication origins.

A Baseline and dormant origins. Top, experimental procedure. SIRT1 deficient (SIRT1^Null) HCT116 cells were complemented with either WT-SIRT1 (WT^SIRT1) or H363Y-SIRT1 (an inactive mutant; Mut^SIRT1) cDNA. Replication origins were mapped using nascent strand sequencing (NS-seq). Bottom, a genome browser view of NS-seq coverage across a representative genomic region on chromosome 1. Baseline origins (black rectangles) are defined as active in both WT^SIRT1 and Mut^SIRT1 cells, whereas dormant origins (dashed black rectangles) are active only in Mut^SIRT1 cells. Pre-RNase treatment serves as a control, as RNA primer removal allows lambda exonuclease to digest nascent DNA, eliminating origin peaks. B Nascent strand abundance (initiation) at baseline and dormant origins in cells harboring WT^SIRT1 (WT) or Mut^SIRT1 (Mut). C HCT116 cells containing auxin – inducible degron MCM2 (MCM2-mAID) were complemented with either intact (WT) or S139A substituted MCM2. IAA (auxin, 500 μM for 16 h) depleted the endogenous MCM2-mAID. Complementation with MCM2-S139A, but not WT MCM2, inhibited replication initiation (Supplementary Fig. 1E). D The abundance of pMCM2-S139, MCM2, Treslin, MTBP and pRecQL4-S89 in whole cell (WCE) and chromatin extracts (ChrE) from HCT116 cells harboring MCM2-mAID complemented with MCM2-WT and MCM2-S139A with and without IAA. Representative of three independent replicates. The expression of MCM2-WT and MCM2-S139A was similar and comparable to the endogenous protein. E, F Binding sites of pMCM2-S139, MTBP, Treslin, RecQL4, and pRecQL4-S89 mapped by chromatin immunoprecipitation followed by sequencing (ChIP-Seq) in WT^SIRT1 and Mut^SIRT1 cells. pMCM2-S139 binding was measured in G1-synchronized cells, whereas binding of MTBP, Treslin, RecQL4, and pRecQL4-S89 was measured in asynchronous cells. E ChIP-seq coverage at the genomic region depicted in panel (A). From top, baseline and dormant origins; averaged ChIP-seq coverage (from two biological replicates) for MTBP, Treslin, pRecQL4-S89, pMCM2-S139, pMCM2-S139, RecQL4 and input controls from WT^SIRT1 and Mut^SIRT1. F Heatmaps depicting chromatin binding at baseline and dormant replication origins in HCT116 cells harboring WT^SIRT1 or Mut^SIRT1 as measured as in panel (E). Baseline and dormant origins were stratified based on replication initiation activity as depicted in panels (A) and (B). G Quantification of chromatin binding patterns shown in panel (F) (above). Source Data are provided as a Source Data file.

Fig. 2. Timing of MTBP and pRecQL4 association with replication origins during S-phase progression.

A Experimental procedure. HCT116 cells were released from a double thymidine block (DTB) and collected asynchronously (Asy) or at the indicated time points, corresponding to the early, mid-, and late-S phase of the cell cycle. Replication initiation sites were mapped by NS-seq and chromatin binding patterns of the indicated proteins were determined by ChIP-Seq. B Confirmation of cell synchronization by flow cytometry. C Nascent strand abundance in HCT116 cells released from DTB, as indicated in panel (A), and collected 0 h, 1 h, 4 h, and 8 h post-release and in asynchronous (Asy) HCT116 cells. Early, mid, and late-initiating replication origins were stratified based on replication timing profiles collected as indicated in Supplementary Fig. 2E. ChIP-seq using MTBP (D) and pRecQL4-S89 (E) antibodies was performed in cells collected at the same time points as in panel (C). The heatmaps show MTBP and pRecQL4-S89 signal strengths in early-, mid-, and late-replicating chromatin centered on origins stratified as in panel (C). For MTBP, which does not bind dormant origins, only baseline origins are shown. For pRecQL4-S89, the left panel shows binding to baseline origins and the right panel shows binding to dormant origins. F A model showing the selective association of MTBP and pRecQL4-S89 with replication origins, regulating the initiation of DNA replication. Dashed ellipses represent diffused binding, whereas solid ellipses indicate enriched binding.

Fig. 3. MTBP and pRecQL4 association with dormant origins.

A Effect of the MCM2^S108A mutation on interactions with MTBP and pRecQL4. Experimental procedure: HCT116 cells with MCM2-mAID were complemented with WT or S108A MCM2 constructs. IAA treatment (500 μM for 16 h) depleted endogenous MCM2-mAID, enabling the incorporation of WT or S108A MCM2 into the pre-replication complex. Treatment with the SIRT1 inhibitor Ex527 (1 mM every 24 h for 2 days) activates dormant origins in cells complemented with the intact, but not the S108A mutated MCM2. B Replication origin binding of MTBP (left) and pRecQL4 (right) in unperturbed and Ex527 treated (1 mM every 24 h for 2 days) HCT116 cells harboring either WT-MCM2 or S108A-MCM2. Baseline and dormant origins were categorized as described in the legend to Fig. 1A, B.

Fig. 4. MTBP and pRecQL4 association with baseline origins after perturbation of DNA replication.

A Recovery of HCT116 cells from APH-induced replication stress. Top, experimental procedure. Below, representative quantitative immunofluorescence-based cytometry (QIBC) of cells exposed to APH (10 μM for 1 h) and released for the indicated time frames. B EdU incorporation in cells treated as indicated in panel A, quantified across 4 biological replicates: Control (n = 5010), APH (n = 5706), 1 h (n = 5648), 4 h (n = 11,951), 8 h (n = 8528), 24 h (n = 7498), and 36 h (n = 5024). Stacked bars show the mean, error bars indicate standard deviation (SD). Statistical analysis was performed for the “high EdU” population compared to the control using a two-sided paired Tukey test, adjusted for multiple comparisons: ***p = 0.0008; ****p < 0.0001; ns not significant. C Binding of MTBP (left) and pRecQL4-S89 (right) to baseline and dormant origins in HCT116 cells exposed to aphidicolin (APH, 10 μM for 1 h) and released at the indicated time points. D Fractions of baseline (top) and dormant (bottom) origins binding to MTBP, pRecQL4, or both, at the indicated timepoint post-APH removal. Vehicle (DMSO) treated (veh.) and untreated control (cont.). The line plot represents the average of two biological replicates. E Replication stress facilitates an interaction between MTBP and pRecQL4. Left, experimental strategy. Right, Chromatin extracts from asynchronous cells untreated or treated with APH (10 μM for 1 h) were first immunoprecipitated (IP) with PCNA. PCNA-depleted fractions were then immunoprecipitated (pooled five reactions) with pMCM2 antibodies to isolate pre-replication complexes. MTBP or pRecQL4 antibodies were used for further IPs (with three pooled reactions), and the presence or absence of MTBP or pRecQL4 was detected by immunoblotting. Representative immunoblots of 3 independent replicates. Source data are provided as a Source Data file.

Fig. 5. Role of RecQL4-S89 phosphorylation in regulating MTBP association with replication origins.

A Nascent strand abundance (left, red) and MTBP association (right, green) at baseline and dormant replication origins in unperturbed RecQL4 CRISPR knockout HCT116 cells (Null) complemented with either WT-RecQL4 or S89A-RecQL4. Baseline and dormant origins were stratified as described in the legend to Fig. 1. B MTBP binding at baseline and dormant replication origins in the cell variants described in panel A, without APH or 1 h after release from APH exposure (10 μM for 1 h). C, D The recovery of HCT116 cells from exposure to APH. Representative immunofluorescence (C) and Quantitative immunofluorescence-based cytometry quantification (D) of DAPI (a DNA content indicator), EdU (a DNA synthesis marker), RPA2 (a replication stress marker), and γH2AX (a DNA damage marker). HCT116 cells harboring either WT-RecQL4 (WT) (D, top panel) or S89A-RecQL4 (D, bottom panel) were untreated (UT) or exposed to APH (10 μM for 1 h) and then released (rel.) into a fresh medium for the indicated time periods. For each condition, cells were pre-labeled with EdU (10 μM for 30 min) and then detergent-extracted and fixed (see methods). Scale bar = 2 μm. EdU versus DAPI levels were plotted, and high-, low-RPA positive and RPA negative (neg.) cells (for calibration, see Supplementary Fig. 4E) are shown in red and blue, respectively. A minimum of 4 replicates were analyzed and plotted together. E Quantification of RPA and γH2AX in untreated cells and in cells recovering from exposure to APH as described in Fig. 5C, D. (for QIBC profiles, see Supplementary Fig. 5G and source Data file). Total number of cells quantified across 4 biological replicates: for WT- UT (n = 5010), 1 h (n = 5648), 4 h (n = 11,951), 8 h (n = 8528), 24 h (n = 7498) and S89A-UT (n = 5024), 1 h (n = 6443), 4 h (n = 8601), 8 h (n = 9288), 24 h (n = 6186). Stacked bars show the mean, error bars indicate SD. F Left, Colony formation in HCT116 cells harboring WT (left) and S89A-RecQL4 (roight) exposed to 10 μM APH for the indicated times. Right, mean colony formation in three independent experiments. Error bars indicate SD. Source data are provided as a Source data file.

Fig. 6. A model illustrating dynamic interactions at replication origins during the recovery from replication stress.

Left, in cells harboring WT-RecQL4, recruitment of the Treslin-MTBP complex marks a sub-group of baseline origins (e.g. origin 1) whereas pRecQL4-S89 (pRecQL4) associates with dormant origins (e.g. origin 2). When cells encounter replication stress, baseline origins can stall (red halo) but dormant origins, which do not initiate replication, are not affected and still maintain MCM complexes (green ring). Although pRecQL4 binding does not allow initiation from dormant origins under normal circumstances (see Fig. 2F), RecQL4 dissociates from origins when cells recover from replication stress and allows MTBP re-association with those origins. Upon the binding of MTBP and Treslin to dormant origins, these origins initiate replication to complete DNA synthesis. Right, in cells that do not harbor pRecQL4-S89 (either RecQL4 depleted or harboring the RecQL4-S89A substitution), MTBP and Treslin associate with both origins 1 and 2, and replication initiates from both origins as cells encounter acute replication stress. Under these conditions, replication from dormant origins cannot rescue the damage after replication stress is removed, preventing normal recovery and leading to the accumulation of ssDNA and subsequent DNA damage.