Over the past two to three decades, replication stress—defined as perturbations that stall DNA replication forks—has emerged as a key driver of genomic instability, tumorigenesis, and a range of developmental disorders. Yet, we are only beginning to understand how cells mitigate endogenous replication stress that arises not only from metabolic fluctuations but also from the intrinsic stochasticity of molecular processes governing DNA replication and cell cycle progression. My ERC starting grant application is dedicated to addressing these fundamental questions—specifically, how core regulators of DNA replication and the cell cycle—such as the replisome and Cyclin–CDK complexes—integrate diverse physiological and pathophysiological cues. To dissect these mechanisms, we use mammalian stem cells and cancer cell models, combining CRISPR-based genome editing, inducible oncogene systems, and advanced methodologies including quantitative single-cell and single-molecule imaging (QIBC, STED, MINFLUX), biochemistry, and genomics.
1. DNA strand-specific rate-limiting mechanism of the human replisome
Recently, we uncovered a previously concealed, strand-specific rate-limiting mechanism operating during unperturbed DNA replication (Chhetri et al., ‘accepted in principle’ Nature 2025). Using complementary biochemical, genetic, and imaging approaches, we show that excessive replication origin firing depletes chromatin-bound PCNA and associated lagging strand components, forcing cells to rely on non-canonical pathways that ultimately culminate in replication collapse3. Mechanistically, we identify PAF15 (PCNA associated factor 15)4-6 as a critical rate-limiting factor that stabilizes PCNA specifically on the lagging strand, via its high-affinity PIP box and unique insertion into the DNA-encircling PCNA channel. This stabilization is antagonized by the PCNA unloader ATAD5, whose activity increases upon PAF15 loss or physiological exhaustion, further impairing lagging-strand synthesis. Moreover, we find that PAF15 levels are tightly regulated by E2F transcription factors to maintain a protective PCNA–PAF15 ratio, while the TIMELESS–CLASPIN complex prevents inappropriate PAF15 access to leading strand-bound PCNA.
2. Redox-sensitive replisome dynamics and cell cycle rhythms:
Proper coordination between cellular metabolism and DNA replication is essential for maintaining genome integrity, especially in rapidly dividing cells during development and cancer. Our earlier work uncovered a redox-sensitive mechanism where reactive oxygen species (ROS), generated under nucleotide depletion, activate the replisome-associated sensor PRDX2. This leads to the eviction of the fork accelerator TIMELESS, triggering a controlled slowdown of replication forks that protects cells from genome instability and replication-induced cell death.
A. Building on these insights, our ongoing work reveals that loss of PRDX2 that drives unscheduled acceleration of replication forks even under dNTP-limiting conditions leads to widespread misincorporation of ribonucleotides (rNTPs) into DNA, thereby creating a critical dependency on RNase H2 to excise embedded rNTPs. These findings raise a fundamental question: how do replicative polymerases maintain nucleotide selectivity and discriminate between rNTPs and dNTPs during the inherently stochastic process of exponential DNA replication, particularly given the vast excess of rNTPs over dNTPs in the cellular milieu?
B. Furthermore, in the absence of PRDX2 and accelerated replication forks, we find that accumulated rNTPs not only compromise genome integrity but also elicit a robust innate immune response—a pathological feature characteristic of inflammatory disorders such as Aicardi–Goutières syndrome (AGS), which is associated with defective rNTP processing from DNA. To investigate these mechanisms in a physiologically relevant context, we employ human neural stem cells (NSCs), combining CRISPR-mediated knockout of PRDX2 and endogenous tagging of the TIMELESS–TIPIN complex with advanced quantitative cell biology and high-resolution metabolomics to dynamically monitor rNTP/dNTP pool during DNA replication.
