Download PDF ArticleOpen accessPublished: 01 July 2026Vincent Gaggioli1,2 na1,Kaustav Sengupta ORCID: orcid.org/0000-0003-0603-09161,2,3 na1,Ashutosh Choudhury ORCID: orcid.org/0000-0002-3815-88531,2,Joanna Paulson1,2,Raviprasad Kuthethur ORCID: orcid.org/0000-0003-0697-78351,2,Collin Bakker ORCID: orcid.org/0009-0005-1322-96501,2,Jialun Li1,2,Calvin S. Y. Lo1,2,Alex Whale ORCID: orcid.org/0000-0003-4568-711X4,Jeroen van den Berg ORCID: orcid.org/0000-0002-9430-71555,6,Leire Asua Intxausti ORCID: orcid.org/0009-0005-6028-88567,8,Noelia Gil-Lanza ORCID: orcid.org/0009-0003-3763-87297,8,Iván Galván-Femenía9,Eleni Maria Manolika1,2,Sangavi Eswaran ORCID: orcid.org/0009-0002-0793-59281,Hina N. Khan1,Estel Ferré10,Gregorie Stik ORCID: orcid.org/0000-0002-1404-199210,Alexander van Oudenaarden ORCID: orcid.org/0000-0002-9442-35515,6,Argyris Papantonis ORCID: orcid.org/0000-0001-7551-107311,Jonathan Houseley ORCID: orcid.org/0000-0001-8509-15004,Sriram Sridharan ORCID: orcid.org/0000-0002-4811-819312,13,Arnab Ray Chaudhuri1,Aleix Bayona-Feliu ORCID: orcid.org/0000-0002-7412-16527,8 &…Nitika Taneja ORCID: orcid.org/0000-0002-5513-52821,2 Nature (2026) Cite this articleSubjectsChromatin structureStalled forksAbstractReplication stress poses a major threat to genome integrity, yet how higher-order chromatin organization contributes to replication fork protection remains unclear1,2. Here we show that replication stress induces the formation of transient chromatin loops that enclose de novo heterochromatin-enriched stalled replication forks3. Stressed forks preferentially stall at convergent CTCF motifs, triggering stress-dependent CTCF enrichment that constrains loop extrusion and stabilizes these structures. Loop stabilization requires both CTCF anchoring and G9a-dependent heterochromatin (trimethylation of Lys9 of histone H3 (H3K9me3)) deposition on nascent DNA within the loop body. These loops function as protective scaffolds that shield stalled and reversed forks from degradation by multiple nucleases. By contrast, combined loss of stress-induced heterochromatin and CTCF enrichment destabilizes the loop scaffold, exposing multiple entry points for nucleolytic attack and resulting in extensive nascent-strand degradation through mechanisms distinct from classical fork-reversal-dependent pathways. This protective architecture is similarly critical in BRCA2-deficient cells, in which replication-stress-associated loops predominantly safeguard replication initiation zones, while nascent DNA outside these loops undergoes massive degradation and remains highly susceptible to mutations. Our study elucidates the fundamental role of replication-stress-induced three-dimensional genome reorganization in preserving replication fork stability, thereby mitigating mutagenesis and genomic instability.MainDNA replication is frequently challenged by endogenous and exogenous insults that induce replication stress, a major driver of genome instability and cancer4,5. These challenges, including diverse DNA lesions, restricted nucleotide availability and transcription–replication conflicts, impede fork progression and necessitate protective responses6,7.Replication timing is closely linked to chromatin organization, with early-replicating regions associated with open chromatin and late-replicating regions enriched in heterochromatin8. This spatiotemporal organization influences susceptibility to replication stress and genome instability2,8,9,10. In response to stress, cells activate checkpoint pathways to stabilize stalled replication forks, where nascent DNA is highly vulnerable to degradation11,12. Concomitantly, chromatin undergoes dynamic reorganization when DNA replication forks are challenged1,2. We previously showed that the histone methyltransferase G9a (encoded by EHMT2) promotes de novo H3K9 methylation at stressed replication forks, driving local chromatin compaction and protecting nascent DNA from error-prone polymerases to facilitate fork restart3.However, whether such local chromatin changes scale to higher-order genome organization remains unclear. Here we show that replication stress induces ATR-dependent CTCF anchoring and G9a-mediated heterochromatin assembly to form transient chromatin loops at stressed forks. These structures create a protective architectural scaffold that limits nuclease access and stabilizes replication intermediates.Mapping heterochromatin at stressed forksUsing our single-molecule chromatin fibre ChromStretch assay, we previously showed that hydroxyurea (HU)-induced (1 mM HU, 1 h) fork stalling triggers transient, G9a-dependent de novo H3K9me3 deposition on nascent DNA, at single forks and replication bubbles, that dissipates quickly once stress is relieved3.To determine whether this response is conserved, we analysed cells exposed to diverse replication stressors (0.2–1 mM HU, camptothecin (CPT), aphidicolin (APH), cisplatin) and observed a common accumulation of H3K9me3 at 5-ethynyl-2′-deoxyuridine (EdU)-labelled replication sites, albeit to varying degrees. This indicates a general, replication stress-dependent chromatin response (Fig. 1a,b and Extended Data Fig. 1a,b).Fig. 1: Single-molecule analysis and genome-wide mapping reveal de novo heterochromatin formation at stressed DNA replication sites.Full size imagea, Representative chromatin fibres (ChromStretch) without treatment (left), with 1 mM HU (middle) or with 5 µM APH (right) for 1 h. EdU (red), H3K9me3 (green) and H3 (blue) are visualized. Bottom, the corresponding intensity quantifications. a.u., arbitrary units. Scale bar, 2 μm. b, The distribution of H3K9me3 intensity at EdU spots. Numbers of EdU sites analysed per condition: from left to right, n = 72, 75, 68, 70, 77 and 76, imaged from one representative experiment, which was been performed twice with similar results. Statistical analysis was performed using Kruskal–Wallis tests followed by Dunn’s test; from bottom to top, P